MBL/WHOI i JOURNAL OF SHELLFISH RESEARCH VOLUME 15, NUMBER 1 APRIL 1996 The Journal of Shellfish Research (formerly Proceedings of the National Shellfisheries Association) is the official publication of the National Shellfisheries Association Editor Dr. Sandra E. Shumway Natural Science Division Southampton College, LIU Southampton. NY 11968 Dr. Standish K. Allen, Jr. (1996) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Peter Beninger( 1997) Department of Biology University of Moncton Moncton, New Brunswick Canada El A 3E9 Dr. Andrew Boghen (1997) Department of Biology University of Moncton Moncton, New Brunswick Canada ElA 3E9 Dr. Neil Bourne (1996) Fisheries and Oceans Pacific Biological Station Nanaimo, British Columbia Canada V9R 5K6 Dr. Andrew Brand (1996) University of Liverpool Marine Biological Station Port Erin, Isle of Man Dr. Eugene Burreson (1997) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Peter Cook (1996) Department of Zoology University of Cape Town Rondebosch 7700 Cape Town, South Africa EDITORIAL BOARD Dr. Simon Cragg (1996) Faculty of Technology Buckinghamshire College of Higher Education Queen Alexandra Road High Wycombe Buckinghamshire HPll 2JZ England, United Kingdom Dr. Leroy Creswell (1997) Harbor Branch Oceanographic Institute US Highway 1 North Fort Pierce, Florida 34946 Dr. Ralph Elston (1996) Battelle Northwest Marine Sciences Laboratory 439 West Sequim Bay Road Sequim, Washington 98382 Dr. Susan Ford (1996) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Raymond Grizzle (1997) Randall Environmental Studies Center Taylor University Upland, Indiana 46989 Dr. Robert E. Hillman (1996) Battelle Ocean Sciences New England Marine Research Laboratory Duxbury, Massachusetts 02332 Dr. Mark Luckenbach (1997) Virginia Institute of Marine Science Wachapreague, Virginia 23480 Dr. Bruce MacDonald (1997) Department of Biology University of New Brunswick P.O. Box 5050 Saint John, New Brunswick Canada E2L 4L5 Dr. Roger Mann (1996) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Islay D. Marsden (1996) Department of Zoology Canterbury University Christchurch, New Zealand Dr. Roger Newell (1996) Horn Point Environmental Laboratories University of Maryland Cambridge, Maryland 21613 Dr. Michael A. Rice (1996) Dept. of Fisheries, Animal & Veterinary Science The University of Rhode Island Kingston, Rhode Island 02881 Susan Waddy (1997) Biological Station St. Andrews, New Brunswick Canada, FOG 2X0 Dr. Elizabeth L. Wenner (1996) Marine Resources Research Institute Box 12559 Charleston, South Carolina 29422 Mr. Gary Wikfors (1996) NOAA/NMFS Rogers Avenue Milford, Connecticut 06460 Journal of Shellfish Research Volume 15, Number 1 ISSN: 00775711 April 1996 This special issue was supported in part by the Center for Marine & Estuarine Disease Research National Health and Environmental Effects Research Laboratory U.S. Environmental Protection Agency National Shellfisheries Association and the ^'°'^ ^,T,TP^'' 'n«tLilon Journal of Shellfish Research APR 2 S 1996 Woods Hole. MA 02543 Acknowledgements This special issue of the Journal of Shellfish Research is sponsored in part by the Sea Grant College Programs of Virginia, Maryland, Delaware, and New Jersey, the National Office of Sea Grant, and the National Oceanic and Atmospheric Administration Chesapeake Bay Office through funding from NOAA, Office of Sea Grant, U.S. Department of Commerce, under federal Grant No NA56RG0141. .^-■"^°^. PREFACE The severity and continual spread of Perkinsiis marimis disease and its increasing impact on the economic and ecological resources of the Gulf of Mexico and Atlantic coast of the United States have prompted urgent attention from the scientific community. This imperative has been recognized by the two federal agencies that supported the publication of this document — the Environmental Protection Agency through the Center for Marine & Estuarine Disease Research (CMED) and the National Oceanic and Atmospheric Administration through the Sea Grant Program. At the 1994 National Shellfisheries Association meeting in Charlestown, South Carolina. CMED sponsored a day-long symposium on P. marinus disease of oysters. The participants agreed to prepare review articles that would include recent progress and overall perspective in their areas of expertise. The valuable efforts of these scientists are included in this special issue of the Journal. The issue also offers an opportunity to recognize three scientists for their creative and enduring investigations into P. marinus disease — Sammy M. Ray, Jay D. Andrews, and Frank O. Perkins. Dr. Ray was a member of one of the original teams that discovered the parasite in Gulf of Mexico oysters and. as a graduate student, developed the diagnostic culture medium and staining technique that has been used almost exclusively for over 40 years. Dr. Andrews was one of the earliest researchers to investigate the parasite in Chesapeake Bay. and in the ensuing 33 years, he has monitored resident and caged oysters in Virginia waters to establish much of our knowledge regarding epizootiology and natural transmission of the disease. Dr. Perkins has been intimately involved in the morpho- logical description of the pathogen for over 20 years. His ultrastructural descriptions of P . marinus are unparalleled, and in recognition of this, the taxononiic description of this microorganism bears his name. The foreword and first two articles in this issue present historical perspectives from each of these scientists. Their work, as well as their enthusiasm and insight, has been at the core of our research in this field. Their additions to the special issue provide a continuity and context for the work that remains. Completion of this issue would not be possible without the help and support of many people. From EPA. thanks are due to Robert Menzer and Courtney Riordan for their support during the initial stages of the project, and more recently to Sonny Mayer and Gil Veith; from Sea Grant, similar thanks are due to Bess Gillelan for her initial support, and more recently to Bill Rickards and James McVey. Several authors have reminded me of our mutual appreciation for the many pcer-rcviewcrs. most of whom provided excellent reviews and some on extremely short notice. I greatly appreciate the capable assistance and moral support of Jill Adams as well as the advice and publication assistance of Sandra Shumway. editor of JSR. The symposium and the special issue would not have been possible without the membership of the National Shellfisheries Association, whose continued interest and support of shellfish research have provided venues for these and many other productive projects and programs. William S. Fisher Editor FOREWORD Frank O. Perkins Over 45 years have elapsed since John G. Mackin. H. Malcolm Owen, and Albert Collier of the Texas A & M Research Foundation and the Louisiana Department of Wildlife and Fisheries first noted that a protistan parasite was associated with mortalities of Crassostrea virginica found m the area of the Mississippi River delta. Due to the presence of cells of a parasite with a large eccentric vacuole containing a prominent inclusion, they concluded that the protist was a species of the genus Dennocystidium and named it Dermocys- tuUuin manintm. In the next two decades, it was well documented that the parasite was the causative agent of the oyster mortalities first observed in the Gulf of Mexico coastal waters and, in fact, could be found in oysters from Texas to New Jersey as well as other bivalve molluscs in that range. Although there were many researchers who contributed to our knowledge of the parasite during the 1950s and 1960s, Jay D. Andrews, John G. Mackin, and Sammy M. Ray provided the ma|ority of information. Ray facilitated investigations of the parasite by providing the fluid thioglycollate medium (FTM) technique by which rapid and inexpensive detection of cells of the organism could be accomplished in large numbers of oyster tissue samples due to marked enlargement of the pathogen in the culture medium. Although not yet rigorously evaluated, there is good evidence that enlargement occurs without cellular multiplication. Almost 40 years later the Ray technique was redesigned to permit a quantitative estimation of the numbers of cells in selected tissues and in whole oysters by incubation in FFM followed by digestion in an NaOH solution and counting the number of Perkmsiis mannus cell walls in the digest. Detailed epizootiological and experimental ecological studies by Andrews, Mackin. and Ray yielded information of value to oyster growers and managers of oyster populations. It was soon determined that transmission of infections occur from oyster to oyster and the pathogen is most virulent at higher temperatures and salinities approximating 20 to 30°C and 20 to 30 ppt. respectively. The structure and life cycle of the parasite was examined in greater detail in the 1960s and 1970s using electron microscopy. The demonstration of zoosporulation in sea water of cells (hypnospores) which had enlarged in FFM yielded the observation that infective, biflagellated zoospores were released. These cells were found to have an apical complex and other apicomplexan structures. Thus, evidence was presented that the organism is closely related to the Apicomplexa. Norman D. Levine in 1978 renamed the parasite P. marinus and placed it in a new class Perkinsea in the phylum Apicomplexa. A curiosity which remains to be explained is the fact that for about 20 years after its discovery, zoosporulation could be readily induced to occur in P . marinus under laboratory conditions by most cells of a hypnospore population. Since the late 1980s, this was found to occur in less than 1% of a hypnospore population and zoospore release failed to occur. On the other hand, isolates of Perkinsiis spp. hypnospores derived from other bivalve hosts readily zoosporulate in sea water. During the 1980s until present, research activity involving P . marinus and other species in the genus increased markedly on several fronts and excellent progress has been made. Epizootiological studies and investigations in experimental ecology both in the field and laboratory have centered around the effects of salinity and temperature in controlling expression of the disease, thus expanding on the 6 Perkins extensive studies reported earlier by Mackin and co-workers. The workers who followed them have confirmed that increased salinity and temperature enhance expression of the disease, and they have greatly expanded upon our knowledge of the details of that paradigm as well as the exceptions. Salinity has emerged as a more dominant factor than temperature in some studies and conditions. In other studies, temperature has been found dominant. The laboratory component of temperature and salinity studies has centered mainly around observations of hemocyte function and composition of hemocyte populations. Large-scale, field observations have revealed that epizootics of P. niannus in oysters are not induced simply by fluctuations in temperature and salinity but rather some other stimulatory factor or factors such as limited food supply or recruitment that occurs just before or at the same time as elevated temperatures and salinities. Thus, progress is being made toward constructing climatic models to predict the activity of the pathogen. Of significance may be the recent observation that P. marinu.s proliferation is enhanced by excess iron accumulation in the host. It is known that there is an increase in iron levels in oysters in the summer when P martnus causes elevated mortalities. Thus, researchers will undoubtedly have to consider many more factors than just salinity and temperature in their modelling efforts. Oyster hemocyte and P. marinus interactions have been evaluated in vitro by measuring reactive oxygen mtcrmcdiates (ROD primarily as expressed by the luminol-enhanced chemiluminescence (CD response and by assaying for hemocyte lysosomal enzymes. Although there are conflicting results, it appears that P. marinus can either prevent ROI production or neutralize ROI in hemocytes with one hypothesis being that acid phosphatase produced by P. marinus inhibits superoxide radicals released by hemocytes. In the future, there will undoubtedly be increased research activity directed toward understanding how P. marinus is able to survive and multiply in hemocytes, recognizing that the hemocytes are the primary line of host defense against microbial agents. Related to this are ongoing investigations to identify and quantify substances that are produced by the pathogen and result in destruction of oyster cells. The question as to whether anthropogenic chemicals in growing waters predispose oysters to mortalities caused by P. marinus continues to be one of major importance to managers and users of the estuarine environment. For over a century, oyster farmers and harvesters have cited pollution as the primary reason for the decline in oyster production with enhancement of microbially induced disease by pollution being a focus of their complaints. However, the evidence to support or refute their claims is not yet sufficient. In recent years, some insight has been obtained in working with compounds such as tributyltin and sediments contaminated by polynuclear aromatic hydrocarbons. It appears that some anthropogenic compounds can enhance the proliferation of P. marinus in oysters and can suppress the CL response. Furthermore, the matter of soluble iron (mentioned above) needs to be considered. It has been suggested that increased iron levels in industrially contaminated waters and/or sediments may enhance the expression of the disease. Therefore, the long-standing complaints of the oyster harvesters may prove to be correct. However, much is left to be determined before proof is forthcoming and infomied management decisions can be made relevant to this issue. Also of importance to managers and users of oyster populations is the question of whether Perkinsus sp. or spp., which are found in most (all?) other bivalve mollusc species co-existing with C. virginica, are P. marinus or another species of Perkinsus. There is probably at least one other species of Perkinsus in bivalve molluscs associated with C. virt^inica. It is found in Macoma ballhica and Macoma mitchetti and is probably Perkinsus atlantwus. It can be induced to infect C. virginica most easily when its zoospores are fed to oysters. The observation of Perkinsus cells in other bivalve molluscs may in large part involve a carrier relationship with the pathogen, but multiplication of P. marinus is known to occur in many of those presumptive carriers. Whether they cause mortalities in those bivalves remains to be seen. This information is of interest to governmental regulators of bivalve mollusc transportations between estuaries and must be more completely investigated. Recently, evidence has been obtained that there are probably strains or races of P. marinus with the Gulf of Mexico strains being less virulent than those along the mid-Atlantic Ocean coast of the U.S. Such prclmiinury information requires further clarification so that more informed decisions can be made concerning transportation of oysters. Although some excellent biochemical and physiological studies were conducted using P. marinus cells isolated from infected oyster tissue, the lack of axenic cultures inhibited pursuit of such research. The problem was solved in late 1992 and early in 1993 followed by publication in 1993 of scmi-deflned culture media formulations by three different laboratories within months of each other. The contributions were significant and, as expected, have resulted in improved ability to investigate the biological characteristics of the pathogen. A further refinement has been made with the formulation of a defined culture medium which will permit even greater biochemical and physiological characterizations. The only word of caution has been that the few studies of transmission of infections using cultured cells have resulted in the observation that such cells do not appear to be as infective as P. marinus isolated from oysters and used directly in challenge experiments without being cultured. Identification of a culture medium that yields cells of the same infectivity as uncultured ones must be accomplished to lessen uncertainty as to whether naturally occurring characteristics are being observed when cultured cells are used. It is known that the cytological characteristics of many of the cells in culture differ in terms of size and cytokinesis from those observed in oyster tissues. It has been established that infections of P. marinus occur from oyster to oyster, the developmental cycle in the oyster appears to have been well characterized, and it is known that zoosporulation can occur outside of the host to yield zoospores that are infective for other oysters. Nevertheless, the question has remained as to whether saprobic development can occur free of the host. The fact that the pathogen can be cultured in a variety of media leads one to suggest that P. marinus. as well as other species of Perkinsus. is a faculative pathogen. With the provision of fluorescein-labeled specific antibodies to P. marinus. cell DNA labeling with propidium iodide, and the use of flow cytometric analyses, it is now possible to detect cells of Perkinsus (not just P. marinus) in water and sediment samples. This will undoubtedly lead to greater insights into the life cycle with answers to the question of whether there is multiplication of the pathogen free of its host. Evidence that this may occur comes from the observation that enlarged cells (hypnospores) in sea water may not zoosporulate but rather may form hyphal-like outgrowths into which the cytoplasm flows and subdivides into daughter cells that are released into the sea water. These daughter cells are morphologically dissimilar to those found in the host. Whether these cells are saprobic forms in the life cycle or must enter a host to continue development remains to be determined. Foreword 7 Whereas most investigators accept that Pcrkinsus spp. are related to the Apicomplexa, the taxonomy and phylogeny of the pathogens remain a subject for scrutiny and reevaluation In hghi of new phylogenetic alignments of the Protista and recent findings by molecular biologists studying nucleic acid base sequences of P inariiiiis. as well as the reinterpretation of the morphology of the pathogen by others, it is now realized that pathogens in the genus belong either with the Dinonagellata. the Apicomplexa. or some intermediate taxon yet to be described. It has alrcads been suggested that the Apicomplexa arose from the dinoflagellates with Perkinsus spp. being an early diverging group in the evolution of the Apicomplexa. This hypothesis may prove to be accurate when an adequate number of species in the two higher taxa are thoroughly evaluated. The reader of this special issue of the Joiirntil af Shellfish Research will find most of these research accomplishments described in greater detail in the papers that follow as well as other aspects not covered in this introduction. The accomplishments are considerable and much valuable information will undoubtedly continue to be provided in the years ahead. As measured by publications, the rate at which new infonnation was being provided reached its highest level in the early 1990s and continues today. This is due in large part to funding from NOAA and in particular from NOAA"s National Oyster Disease Research Program which has provided over $6 million in funding before being terminated this year. The U.S. Congress funded the Program with a special appropriation following an initiative by former Congressman Roy Dyson with particularly strong support from the Virginia and Maryland delegations; therefore, many of us who have worked on P . marinus are indebted to them. Once commercially viable answers to oyster mortalities caused by P. mariniis have been found, oyster harvesters and farmers will also have these Congressional representatives to thank for much of the progress made in attaining that goal. Frank O. Perkins Virginia Institute of Marine Science College of William and Mary Gloucester Point. Virginia 23062 Journal of Shellfish Rf.seanh. Vol. 15. No. 1.9-11, 1946. HISTORICAL PERSPECTIVE ON PERKINSUS MARINUS DISEASE OF OYSTERS IN THE GULF OF MEXICO SAMMY M. RAY Marine Biology Dept. Texas A&M V niversity-Gaheston Galveston, Texas 77553 ABSTRACT A brief history of events and individuals involved in the discovery of the important oyster pathogen Perkinsiis marinus (Dermo) is presented. .^Isoa short review of the development of the tluid Ihioglycollate culture technique (FTMl for diagnosing Dermo is provided. In 1946 the oystcmicn of Louisiana filed $30-$40 million law- suits against several major oil companies and the Freeport Sulphur Company for alleged mortality of oysters due to in-shore petro- leum operations. The primary allegation was that the discharge of "bleed" or "production" water into oyster-producing bays (pri- marily west of the Mississippi River delta) was responsible for abnormal losses of market-sized oysters. With the filing of these lawsuits, the defendants and plaintiffs began assembling teams of experts to investigate the allegations. Four major research groups were charged with determining the role, if any, of petroleum op- erations in oyster mortalities and determining the cause(s), if pos- sible, of such high mortalities. They included; 1. Texas A&M Research Foundation (TAMRF) Project 9. This effort, which was by far the largest, was funded by several oil companies. The Project 9 investigators included Drs. Sewell H. Hopkins (head), John G. Mackin, and Win- ston Menzel as well as several chemists, supporting scien- tists, and technicians 2. Gulf Oil Corporation (Gulf). Rather than join the TAMRF group. Gulf retained Albert W. Collier to lead its investi- gation. Gulf believed that more diversity would be intro- duced with two investigating groups. A. Wayne Magnitzky. Joe O. Bell, and Sammy M. Ray were hired by Collier to assist in the Gulf studies. Primary chemical support was from scientists at the Mellon Institute in Pittsburgh, PA. 3. Louisiana Wildlife and Fisheries Commission (LWFC). The LWFC investigation was led by Dr. H. Malcome Owen with the assistance of Robert M. Ingle, Fred (Red) Brig- ance, Lester W. Walters, and William Tolbert. 4. Freeport Sulphur Company (Freeport). Freeport's investi- gations were headed by Dr. A. E. Hopkins. To the best of my knowledge, these investigations were largely concerned with determining the effects of sulfur "bleed" or "produc- tion" waters on oysters. 1 recall only a few of the names of persons that assisted Dr. Hopkins, but they included Robert P. Hoffstetter, a cooperative student from Antioch College; John Boss, a cooperative student from Tulane; and Ted Ford. It is of interest to note that at least four of the major investigators of the Louisiana oyster mortality problem (S. H. Hopkins, J. G. Mackin, H. M. Owen, and Winston Menzel) had formerly worked at the Virginia Institute of Marine Science, Gloucester Point, VA. By early to mid- 1947 all groups launched extensive field and laboratory studies designed to determine the effects on oysters of Louisiana crude oil, "bleed" waters, water-soluble oil fractions, oil emulsions, and even associated oil production activities. Field studies by all oyster study groups generally showed two charac- teristic features of the Louisiana oyster mortalities: ( I ) major losses occurred in high-salinity areas during the warm months, and (2) market-sized oysters appeared to be much more susceptible than smaller oysters to mortality. As the groups began to gather results from field and laboratory studies it became apparent to most investigators that oil and its associated operations were not the likely causes of continuing massive oyster mortalities in oysters transplanted from seed grounds east of the Mississippi River to oyster leases west of the river. A feature that supported this view was the lack of similar mortalities in oil fields located in low-salinity areas. This obser- vation was also believed to be related to the fact that mortalities of the mid- 1 940s coincided with an extensive drought period in Lou- isiana. With lessening concern about oil operations, several investiga- 10 Ray tors began to consider other causes for the abnormal mortahties. The most prominent suspects being considered were Tluiis ( south- ern oyster drill), Nematopsis (sporozoan), and Polydora (mud worm). Yet, extensive studies did not provide reasonable support for any of these primary suspects, particularly in relation to the high mortalities observed during the mid- 1940s. A year or two after the major investigations began, Albert Collier told me that he believed that an unknown microorganism was responsible for the oyster mortality. He showed me spherical organisms in fresh preparations of pericardial tluid from moribund oysters and pointed out that these bodies were not observed in healthy oysters. At this time he thought the organism was a "col- orless" alga. A little later he showed me histological sections prepared of moribund oysters and pointed out "spherical" bodies that I now know were Perkinsus (Dermocysticlium) cells. About the same time Drs. Mackin and Owen were apparently conducting similar studies of fresh preparations and histological sections of "healthy" and "moribund" oysters. Mackin, Owen, and Collier began to compare data and concluded that each was looking at the same "undescribed" organism that they suspected was the cause of abnormal warm-weather oyster mortalities in high-salinity ar- eas. This collaboration led to the publication in 1930 of the descrip- tion of this unknown agent as DennocystuUiiin inarinum iDermo) by Mackin, Owen, and Collier (Science. Ill, 1930). Dr. Owen found Dermo in some preserved Louisiana oysters that had been exhibited at a World's Fair (Chicago?) circa 1920. This finding indicated that Dermo was present in Louisiana oysters for at least several decades prior to its discovery. After the publication on Dennocyslidium appeared in Science. the lawsuits began to unravel. Only weak evidence supported the allegations that oil operations were responsible for abnormal Lou- isiana oyster mortalities, yet there was strong epidemiological ev- idence linking them to Dermocysticlium infections. By 1950, some of the lawsuits were dropped and the remaining ones were settled out of court on a nuisance basis for something on the order of $300,000-5400,000. It has been reported that the oil companies spent about $2 million on their investigations. 1 have no idea of the amount that the Louisiana Wildlife and Fisheries Commission spent, but Robert Ingle has indicated to me that it was rather small compared with oil company costs. In addition to the discovery of a major disease-causing parasite of oysters, the Louisiana oyster investigations generated a strong impetus for establishment and expansion of marine science pro- grams in Gulf Coast states. One notable example is the role the TAMRF Project 9 played in creating the Department of Oceanog- raphy at Texas A&M University at College Station. Many of the principals in the Louisiana oyster mortality investigations later became significant contributors to the field of marine science. Prior to the discovery that Dermo was the likely cause of ex- tensive warm-season mortality of market-sized oysters in high sa- linity areas west of the Mississippi River, Louisiana oystermen took steps to compensate for the great loss of oysters. In some areas, 75-100 percent of the market-sized oysters would die dur- ing the summer and early autumn. Since the oystermen wished to have a good crop of oysters for the harvest for the holiday trade (Thanksgiving and Christmas), they initially attempted to compen- sate for the summer mortality by doubling or tripling the seed plantings on leased grounds. This approach was possible because there was ample seed on the grounds east of the Mississippi River. This approach proved ineffective — the high mortality rate contin- ued. Now that we know that Dermo may be transmitted directly, the excessive plantings probably exacerbated the spread of Dermo disease. One thing was obvious to the oystermen. Sub-market-sized oysters that appeared to be growing well in spring would suffer extensive mortality during the following warm months of summer and early autumn. They learned that the extra 4—6 months required for the oysters to reach market size was also the greatest danger period. They correctly concluded that the second summer period in high-salinity areas was deadly for market-sized oysters and should be avoided if at all possible. This realization prompted a drastic change in oyster culture strategy. In high-salinity areas west of the Mississippi River, seed planting was delayed until late summer and early autumn. These oysters were then harvested before the next summer. Using this timing of transplanting and harvesting, they avoided much of the summer mortality, which we now believe was caused by Denno. Since most of these oysters were not large enough for the shucked or half-shell trade, they were canned. This method of marketing oysters proved to be very profitable. Although the period between transplanting and harvesting was rather short (6-8 months), the oystermen made money if they harvested a sack for canning for each sack planted. Oystermen were paid by the yield (cans per unit). In the event the harvest exceeded one for one. they did extremely well. The Louisiana canned oyster industry was even- tually destroyed by cheaper imports of canned oysters. Thus, as a matter of survival, the oystermen learned how to avoid the consequences of Dermo disease even before a large team of scientists was able to determine the cause. Some oyster biolo- gists criticized the method devised by the oystermen as putting too much stress on the seed grounds. Had the canned oyster industry not collapsed, some predicted that the Louisiana oyster seed grounds east of the Mississippi River would have been rumed. With settlement of the lawsuits in 1930, Gulf Oil Co. offered me a fellowship to attend either Rice University or Tulane Uni- versity. My charge was to attempt to culture Dennocystidium in order to fulfill Koch's principles for this suspected pathogen of oysters. I chose Rice so that I could work under the late Dr. Asa Chandler, a world-renowned parasitologist, and began my studies in September 1930. The prevailing view at the time was that Dermo was a fungus, so my major efforts involved the use of various techniques generally employed to culture fungi. With the failure to culture Dermo with the usual fungal tech- niques. Dr. Chandler and I considered that this organism might be an obligate parasite. We immediately changed the approach from attempting to culture Dermo to culturing oyster tissue to provide a medium for culturing Dermo. We realized that this was a formi- dable undertaking because tissue culture science was in its infancy at this time. Also a research of the literature indicated that there had been little success in developing molluscan tissue cell lines in the early 1930s. The immediate problem was to develop procedures for obtain- ing sterile oyster tissue. Initially, excised pieces of gill tissue were stored for 24 hours in sterile sea water fortified with penicillin and streptomycin to inhibit bacterial growth. Gill tissue was selected since ciliary activity could be used readily as a gauge of tissue survival. Excised gill tissues appeared to survive antibiotic treat- ment. Thus the next step was to determine if the treated gill tissues were sterile. Since fluid thioglycollate medium (FTM) is com- monly used to test various items for sterility, the treated tissues were placed in tubes of sterile FTM and incubated for 48 hours to History of Perkinsus in the Gulf of Mexico 11 test for sterility. The tubes of FTM showed no evidence of bac- terial growth after 48 hours' incubation. The next step was to examine the treated gill tissues micro- scopically. My initial reaction was that the cultured gill tissues were filled with large "'oil droplets." Upon closer examination the spherical structures appeared to have a definite wall and 1 discounted their being oil droplets. An examination of controlgill tissues (continuously stored in sea water with antibioticsjdid not show the large sphencal bixiies noted in the cultured tissues. Up to this point 1 had relied solely on microscopic examination of pericardial tluid of oysters to determine if they were infected with Dermo. This system worked fairly well in cases of moder- ately to heavily infected oysters. As a safeguard, however, every oyster used in my studies was fixed for possible histological ex- amination if verification was required. Fortunately. 1 recalled hav- ing seen very large spherical bodies in the pericardial tluid of one oyster several months earlier. At the time sketches were made of the bodies and in my data notebook I recorded "these cells look like Dermo but they are too large." Thus 1 dismissed the thought that these bodies were Dermo. After seeing the large bodies in gill tissues cultured in FTM. stained histological sections (the first for my Rice research) were prepared of the oyster with the large cells and gill tissues (cultured in FTM) with large bodies. The bodies from both sources appeared to be the same — leading both Chan- dler and me to believe we were probably looking at enlarged Dermo cells. Verification of this belief was accomplished by studying a time-sequenced series of stained sections of cultured gill and man- tle tissues of infected oysters. The tissue series included control tissues (uncultured) and a series of tissues incubated in FTM at various intervals ranging from 2 to 48 hours. The controls showed the usual forms of Dermo found in histological sections of infected oysters. In FTM-cultured tissues, a slight enlargement could be detected after 2 hours" incubation and the cells appeared to reach maximum enlargement at 48 hours. As the cells enlarged with increased incubation, the typical Dermo cells began to disappear until none could be detected after 24—48 hours. With these data we were satisfied that Dermo did. in fact, enlarge when incubated in FTM as well as in other nutrient media. Moreover, we were con- vinced that little, if any. multiplication occurred during FTM cul- ture. Thus, we believed that a simple, reliable technique for di- agnosing Dermo disease in oysters had been discovered. In retrospect, this discovery resulted from a combination of good luck, good recordkeeping, and logic. It was a matter of ""luck" that the first tissues cultured in FTM came from a heavily infected oyster. Another important factor was the chance obser- vation of enlarged Dermo-like cells in a live oyster, which (for- tunately) was preserved for possible further examination. The logic came from "connecting" the large spherical bodies from two sources as possibly being Dermo cells. The above comments prompt me to quote the late Dr. Sewell H. Hopkins with regard to luck. Since the late Drs. Hopkins and J. G. Mackin were actively working on Dermo. we wished to share the discovery with them and have their comments concern- ing the validity of our data. These scientists agreed that we had made a significant discovery, which should be published and made available to the scientific community as soon as possible. Near the close of our meeting Dr. Chandler said, "Sammy will be the first to admit that he was lucky." And I replied. ""Yes." Dr. Hopkins' next comment is one that I shall never forget. He said. ""You know luck is a strange thing — a person that works 16 hours a day has twice as much luck as one who works 8 hours a day." This comment on luck by Dr. Hopkins gave my morale as a graduate student a great boost. With this diagnostic tool, which circumvented the use of 'ime- consuming histological verification, and my limitations as a mar- ried graduate student with a family. I made the calculated decision to achieve my immediate goal — complete my graduate studies and obtain immediate answers to the most important questions. Since my studies were partially supported by the G.l, bill. I have esti- mated that my 4-year study at Rice cost the Gulf Corporation about $20.(X)0. My studies occurred just before the era of big-time spon- sored research at universities. Gulf gave no money, no overhead, and no other remuneration to Rice University. During the workshop held on Perkinsus marinus at the National Shellfisheries Association meeting in Charleston, SC, in April 1994, a couple of current researchers told me that I was very close to successfully culturing Perkinsus during my studies at Rice and they wondered why 1 stopped working on this aspect. The above comments are given to explain my reason for shifting from the culture to other aspects of the problem. I am greatly impressed by the progress that has been made in the continuous culture of Per- kinsus in an artificial medium as well as improvements in the sensitivity of the fluid thioglycoUate diagnostic technique. Journal oj Shellfish Reseanh. Vol. 15. No, I, 13-16. 19%. HISTORY OF PERKINSUS MARINUS, A PATHOGEN OF OYSTERS IN CHESAPEAKE BAY 1950-1984 JAY D. ANDREWS School of Marine Science Virginia Institute of Marine Science College of William & Mary Gloucester Point. Virginia, 23062 ABSTRACT The pathogen Perklnsus mariims (Demio) was discovered in Chesapeake Bay in 1950. It was already widely distributed in the Bay and caused annual mortahty below the mouth of the Rappahannock River. Annual mortality in trayed oysters at the Virginia Institute of Manne Science (VIMS) varied annually from 24% to 57% at this most favorable site for the disease. Over 2 million bushels of seed oysters from the James River public beds were transplanted annually to private beds in 4 major growing areas. These were Hampton Roads, lower Bay proper. Mobjack Bay at mouth of York River, and the Rappahannock River. The introduction of HaplosporiJnim ncl.wni (MSXl in 1959 resulted in killing most oysters throughout the Bay. and private planting was abandoned. Extreme dry weather dunng the decade of the 1980s allowed both diseases to spread widely throughout the Bay. and the oysters became scarce everywhere. MSX retreated to its endemic area below the mouth of the Rappahannock River when salinities returned to average levels. Dermo destroyed oysters in the seed area of the James River, and it has persisted there tenaciously with low mortality. Market-oyster production dropped from 2 to 3 million bushels annually during the 1950s to 6.000 in 1993. No seed oysters are available, and planting of private beds has ceased. Recovery is slow, and the oyster industry in Virginia was destroyed. KEY WORDS: History, diseases. Chesapeake Bay. pathogen, mortality, distnbution. oyster culture ORIGIN AND LIFE CYCLE OF PERKINSUS MARINUS The origin of Perkinsiis marinus is obscure. The pathogen is widely spread throughout SE Asia; possibly it was introduced by ship transport during World War II, but mortality of oysters was reported before 1940 in Virginia. Numerous small introductions of Pacific oysters (Crassoslreci gigas) have been made along the east coast of North America from the west coast (Andrews 1979). The disease has not been a problem along the west coast of North America, or in Europe where oceanic climates and upwelling keep waters much cooler than on east coasts. The Pacific oyster was introduced along the west coast of North America before 1900; seed oysters in commercial quantities were imported regularly from Japan to Washington and California after World War II (An- drews 1980). Dermo has not occurred along the western shores of Europe despite many tons of introductions of Japanese oysters in late 1960s and early 1970s. Dermo causes a warm-season disease of eastern oysters iCras- soslrea virginica) in Chesapeake Bay (Andrews 1988). At tem- 'VIMS Contnbution No. 1884. peratures above 20°C, the pathogen multiplies and kills oysters about a month after infection. For rapid proliferation, the disease requires temperatures of 25°C which prevail for about 5 months in Chesapeake Bay waters (Andrews and Hewatt 1957). Mortality ceases by 1 November when water temperatures decline below 20°C; during the 1950s, oysters gradually expelled infections, and from February through April most samples showed no infections by Ray's FTM test (Ray 1952). However, oysters placed in 25°C water during late winter and spring revealed about 20'7f infection within a month (Andrews and Hewatt 1957). These hidden infec- tions became patent in June when temperatures reached 25°C. These over-wintering infections caused deaths by 1 August, and two more generations of infections occurred before mid-October with prevalences of Dermo often at 90'7r to 1007^. A comparison of the life cycle of Dermo in the Gulf of Mexico and in Chesapeake Bay is revealing (Andrews and Ray 1988). Higher winter temperatures in the Gulf allow the pathogen to persist in oysters with patent infections throughout the winter, although intensity and prevalence decline. In Louisiana where most Gulf oysters are grown, salinities fluctuate w idely depending upon Mississippi River flow, resulting in wide fluctuations of the disease by years and areas. Planters there must search for disease- 13 14 Andrews free oysters in low-salinity areas to transplant into high-salinity areas tor growth, fattening, and early marketing. Dermo had a wide distribution in Chesapeake Bay at the time of Its discovery in 1950. it spread more widely into marginal salinity areas during the mid-1960s' invasion of upper Virginia and Maryland oyster beds. It spread into the James River only during the dry period of the 1980s. It persisted tenaciously at low levels of infection most winters. Only during dry summers did the pathogen kill oysters in areas with late-summer salinities <20 ppt. Scarcity of oysters in the lower bay limited the distribution of Dermo to manmade structures and creeks which had regular re- cruitment of new year-classes. Beds leased by private growers became barren because those bottoms are soft and oysters sink in time. This fact is important to any efforts to grow oysters in isolation once disease-free seed is available. TRANSMISSION OF P. MARINUS DISEASE Transmission of Dermo is direct from infected dying oysters to other hosts of the species (Mackin 1962). Proximity to gapers (dying oysters) is necessary because large dosage is required to achieve rapid infection (1 x 10 "') zoospores (Roberts. Virginia Institute of Marine Science [VIMS], pers. comm. 1984). All stages appear to be infective or become so when the host dies and prezoosporangia are released into marine waters. From 1 ,000 to 2.000 zoospores are estimated to be produced by one large spo- rangium (Perkins 1966). Zoospores are produced from prezoo- sporangia after culture in thioglycoUate medium for 24 to 48 hours. Feeding or injecting small amounts of macerated gaper tissues produces infection in nearly all oysters. Infection occurs apparently through the digestive tract as indicted by the location of foci of infection in sectioned live oysters. The role of zoospores in open water infections is unknown. They must be an infective stage, but difficulty in production of this stage in the laboratory has prevented completion of the life cycle for the pathogen after 45 years of research. Distances for isolation of oysters from the dis- ease are speculative, which hampers planning for repopulation of oysters in Chesapeake Bay. A host of scavengers live on oyster beds to feed on oysters killed by predators and diseases (Andrews 1988). Blue crabs and mud crabs (Xanthids) kill small oysters whereas nereid worms, spider crabs, and several small fishes such as blennies, gobies, and clingfish are scavengers quick to snatch bits of loose llesh trom gaping oysters (SCUBA observations). OVER-WINTERING OF P. MARINUS IN CHESAPEAKE BAY The level of over-wintering infection is critical to the infectiv- ity and mortality caused by P mwimis disease the following sum- mer. There is a 5-month period of temperatures above 20°C that favors multiplication by the pathogen. If there were no over- wintering infections, the disease would die out; no alternate host has been identified. During the early 1950s, Delaware Bay plant- ers imported oysters from the eastern shore peninsula of Virginia and introduced the disease there (Ford 1992, 1996). During the 1950s, when Dermo was monitored without inter- ference by Haplosporidum nelsoni (MSX) disease, Ray"s FTM tests showed low levels of over-wintering infection at VIMS pier in samples from February through April. Yet development of a few infections was found in June and July after temperatures reached 20°C to 25°C. Oysters placed in warm water for a month in April had about 20% infection. Disease-free oysters imported in April from the upper James River did not develop infections until August, and mortalities were far less than those of acclimated oysters from a previous year's transplanting. This pattern of in- fection and mortality was derived from 30 years of FTM tests on Virginia oysters. It was apparent that hidden infections were over- wintering, but the stage and site in oysters were not known (An- drews 1988). SPREAD OF P. MARINUS DURING DROUGHT OF 1980S The droughty decade of the 1980s allowed Dermo to spread widely into Maryland and up the James River seed area which is vital to oyster culture in Virginia (Andrews 1988, Burreson and Andrews 1988) (Fig. 1 ). Record high salinities and well-populated contiguous oyster beds allowed the disease to spread rapidly and to kill most oysters in the James River except on a couple of upriver beds. Continued harvesting by oystermen helped to deplete the 15-mile-long area of seed oysters and broodstock. Yet the patho- gen persisted by wintering in quite low-salinity waters with tem- perate winters. Over-wintering prevalences of 100% were found at Point of Shoals, a rather upriver site (Ragone Calvo and Burreson 1994). Probably planters, who transported infected oysters from lower river beds to their upper river private beds, helped spread the disease. For a low price, these planters bought market-size oysters in late spring and held them on upriver beds through the summer for high fall prices. The spread of Dermo into the James River seed area during the 1980s, after 30 years of freedom from disease, was unprecedented (Andrews 1988). The prolonged dry weather during the decade, as well as the complete absence of hurricanes to depress salinities during summers, was critical. Each of three earlier decades had one or more hurricane flooding periods. Salinity regimes in Mary- land were very high with a one-time record of 25 ppt at the Bay Bridge above the oyster-growing area. Dry summers increased salinity levels. Many years of observation of coastal plains estuaries, such as the Great Wicomico and Piankatank Rivers, show that Dermo can persist indefinitely in rather low-salinity rivers once established; only when bay waters are salty from low runoff do the two patho- gens, Dermo and MSX, cause appreciable mortality. These estu- aries are dependent on the bay for their salinity regimes, and little freshwater runoff is available to reduce salinity. Typically, these estuaries get to 1 5 ppt only in late summer and fall when time for disease development is limited. Importantly, these estuaries are effective in producing seed oysters with quite regular spatfalls. The diseases have not affected setting rates in the coastal plain estuaries because low populations of broodstocks are adequate. The James River requires very high oyster populations for produc- tion of seed oysters. SUSCEPTIBILITY OF OYSTERS TO P. MARINUS BY SIZE AND ORIGIN Few seed oysters have been imported commercially from the Carolinas into Chesapeake Bay; however, in the early 1950s when New Jersey planters were transplanting oysters from eastern shore of Virginia, a scarcity for local planters induced a trial of South Carolina oysters. Seaside (Virginia) oysters were found to be highly susceptible to MSX, but Dermo was not found on these beds perhaps because fast growth allowed eariy harvesting. South Carolina oysters matured rapidly in Virginia waters and were somewhat resistant to Dermo. Native yearling oysters from James River, where no selection had occurred, showed strong resistance to Dermo infection in open waters. Heavy dosage in aquaria pro- duced infections. PEKKhM.SL'S MARINUS DiS[;aSF OF OySTERS 15 38' BALTIMORE »*^f|/',....^;^ -^Jk^.^' \ S"^' ^ ?,-'!-. w Figure I. Map of Chesapeake Bay showing major rivers where oysters were grown as discussed in text. This resistance of yearlings to Dermo may allow production ol oysters on isolated beds if moved before their second summer of exposure. If a coastal plain estuary were declared strictly for pro- duction of young seed oysters, by prohibiting private plantings along the shores, a disease-free supply of oysters could be pro- duced annually for transplanting to barren areas in higher salinity waters. Early harvesting would be necessary. This method should be tried because recovery of setting and decline of Dermo in the James River seed area are unpredictable. The price of a pint of west coast oysters in local stores is S8 now. which deprives the author of a seafood that was cheap throughout most of his 40 years of study of oyster diseases, I miss thcni. INTRODUCTION OF H. NELSOM DISEASE IN 1959 The introduction o\H. iwlsani {MSX) (Ford and Haskin 1982) complicated disease problems in Chesapeake Bay. This disease moves rapidly up and down the bay with changes in salinities. It requires salinities of 12 to 15 ppt to infect oysters, but it is easily discharged at 10 ppt. This disease invaded the Maryland part of the bay in the mid-1960s, causing heavy mortality (Andrews 1967). and again in the mid-19X()s. The source of infection by MSX is unknown. Fresh- water flows from the large drainage areas of the James, Potomac, and Susquehannah Rivers reduce salinities in winter and spring, and that sets the patterns of disease for different areas. MSX invades the upper bay rapidly in one year and is usually discharged the tollowing winter. Hurricanes play an im- portant role in controlling MSX by lowering bay salinities during summers, and it is expelled in winters. There have been no sig- nificant hurricanes in the Chesapeake area since 1973. H. nelsoni became the dominant pathogen during the 1960s and 197()s in lower Chesapeake Bay. It kills quicker than Dermo. and it has no need for proximity to infected oysters to produce intec- 16 Andrews tions. Dermo was suppressed by scarcity of oysters in the lower bay, but given 2 or 3 years of exposure, it eventually got into trays and oyster beds. Populated beds in the lower bay were gone after 1961 from MSX ravages. Only when the high-salinity years of the 1980s occurred throughout the bay did Dermo become the domi- nant disease in upper bay estuaries. Sparse populations of oysters were being killed in the lower bay by both diseases. Oyster plant- ing in the lower bay had ceased in 1961 after MSX was imported in 1959. The endemic zone for MSX was from the mouth of the Rappahannock River down-bay, including the lower James River and all of the York River. In up-bay low-salinity areas some oys- ters were still being planted. Dermo was explosively dominant m the upper James River where it had never appeared before. The quick invasion and strong persistence in James River have not been explained adequately, although the cause was definitely high salinities through many very dry years in the 1980s. Durmg the 1990s, 5 years had winter-spring runoff in the bay less than half- average flow. Dry summers sustained the high salinities. Expul- sion of Dermo from the James River seed area is slow and the possibility of its removal is questionable. Wet. cold winters may be helpful. It may take many years before the river develops ad- equate broodstock to begin repopulating the famous river that pro- duced 2 to 3 million bushels of seed oysters annually without fail until the 1980s. LITERATURE CITED Andrews, J. D. 1967. Interaction of two diseases of oysters in natural waters. Proc. Natl. Shellfish Assoc. 57:38-49. Andrews, J. D. 1979. Oyster diseases of Chesapeake Bay. U.S. Niitl. Fish. Sen'ice Mar. Fish. Rev. 41(12);45-53. Andrews, J. D. 1980. A review of introductions of exotic oysters and biological planning for new importations. Mar. Fish. Rev. 42:1-11 Andrews, J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marinus and its effects on the oyster industry. Amer. Fish. Soc. Spec. Puhl. 18:47-63. Andrews. J. D. & W, G. Hewatt. 1957. Oyster mortality studies m Vir- ginia, II. The fungus disease caused by Dermocysiidiuin mariimm on oysters of Chesapeake Bay. Ecol. Moiiogr. 27:1-25. Andrews, J. D. & S. M. Ray. 1988. Management strategies to control the disease caused by Perkinsus marinus. Amer. Fish. soc. Spec. Puhl 18:257-264. Burreson, E. M. & J. D. Andrews. 1988. Unusual intensification of Ches- apeake Bay oyster diseases dunng recent drought conditions. Proc. Oceans 88:799-802. Ford. S. E. 1992. Avoiding the spread of disease in commercial culture of molluscs, with special reference to Perkinsus marinus (Dermo) and Haplosporidium nelsoni (MSX). J. Shellfish Res. 1 l(S):539-546. Ford, S. E. 1996. Range extension by the oyster parasite Perkinsus mari- nus into the northeastern United States: Response to climate change? 7. Shellfi.'.h Res. 15:45-56. Ford. S. E. & H. H. Haskin. 1982. History and epizootiology of Haplo- sporidium nelsoni (MSX). an oyster pathogen in Delaware Bay. 1957- 1980. J. Invertehr. Pathol. 40:118-141. Mackin. J. G. 1962. Oyster diseases caused by Dermosystidum marinum, and other microorganisms in Louisiana. Puhl. Inst. Mar. Sci. Univ. Te.xas l:\n^229. Mackin. J. B.. H. M. Owen. & A. Collier. 1950. Preliminary note on the occurrence of a new protistan parasite. Dermocystidium marinum n. sp. in Crassostrea virginica (Gmelin). Sci. Wash. D.C. 111:328-329. Perkins. F. O 1966. Life history studies oi Dermocystidium marinum, an oyster pathogen. Dissertation. Florida State Univ. 273 pp. Ragone Calvo. L. M. & E. M. Burreson. 1994. Characterization of over- wintering infections of Perkinsus marinus. (Apicomplexa) in Chesa- peake Bay oysters. J. Shellfish Res. 13(1): 123- 130. Ray. S. M. 1952. A culture technique for the diagnosis of infections with Dermocystidium marinum. Mackin. Owen & Collier, in oysters. 5c/. Wash D.C. 166:36tV361. Ray. S. M. 1954. Biological studies of Dermocystidium marinum. Rice Inst. Pamphlet, Spec. Issue, Rice Institute. Houston, Texas. Journal of Shellfish Research. Vol. 15. No. I. 17-34. 1996. EPIZOOTIOLOGY OF PERKINSUS MARINUS DISEASE OF OYSTERS IN CHESAPEAKE BAY, WITH EMPHASIS ON DATA SINCE 1985 EUGENE M. BURRESON AND LISA M. RAGONE CALVO School of Murine Scienci' Virginia Institute of Marine Science College of William and Mary Gloucester Point. Virginia 23062 .ABSTRACT Since 19X7 Perkinsiis marums has been the most Iniportant pathogen of the eastern oyster. Crassostrea yiri>inica. in Chesapeake Bay because of its widespread distribution and persistence in low salinity areas. The pathogen became established on all oyster beds in the Chesapeake Bay as a result of natural spread during the consecutive drought years from 19H5 to 1 9X8 or by movement of infected oysters during the same period. Elevated salinities resulting from drought conditions and concomitant warm w inters allowed P. mariniis to proliferale in what were historically low salinity areas. Oyster mortality was high on most beds and landings of market oysters declmed to record low levels in both Maryland and Virginia during the late 19X(ls and early I99()s. The seasonal periodicity of P. marinus is primarily controlled by temperature. Both prevalence and intensity of infections begin to increase in June as temperature increases above 20°C and overwintering infections begin to proliferate. Maximum values of prevalence and intensity occur in September immediately following maximal summer temperatures. Infection regression occurs during winter and spring as temper- ature declines resulting in minimum prevalence and intensity values in April and May. Prevalence and intensity of P marinus infections in oysters from the James River. VA. over a five year period were significantly correlated with temperature when temperature data were lagged three months. Temperature explained 39% of the variability in prevalence and 46% of the variability in intensity. The relationship between temperature and annual variability in P marinus abundance is somewhat obscure, in part because of the difficulty separating salinity and temperature effects. Nonetheless, data from 19X8 to 1994 from the James River. VA. suggest that abnormally warm winters have a more significant impact on summer P. marinus abundance than abnormally cold winters Salinity is the pnmary environmental factor that controls local distribution and intensity oi P. marinii.'; infections. Long-term oyster disease monitoring along a salinity gradient in the James River, VA, revealed a statistically significant relationship between salinity and P marinus prevalence and intensity. P. marinus infections remain light in intensity and no oyster mortality results if salinity is consistently less than 9 ppt. However, infections may persist for years in low salinity areas. If summer/fall salinities range from 9 to 15 ppt some infections may progress to moderate and heavy intensity, but oyster mortality is relatively low. If summer/fall salinities are consistently greater than 15 ppt. moderate and heavy infections may be numerous and oyster mortality may be high. Field studies in the York Rjver. VA, suggest that new P marinus infections are acquired from July through early October, but peak infection acquisition occurs during late August and is correlated with oyster mortality. The early infection process in oysters and the role of zoospores in transmission dynamics in nature are poorly understood. No direct link between oyster defense mechanisms and control of P. marinus infections has been established. If oyster defense mechanisms do modulate P marinus infections, the components have not been identified. There is little evidence to support the common perception that pollution is responsible for the dramatic increase in P. marinus abundance since 1985. Pathogen abundance is cleariy correlated with salinity increases resulting from drought conditions in the late l9X0s, although there may be subtle effects of toxicants or poor water quality on the host/parasite interaction. KEY WORDS: Perkinsus, oyster disease, annual cycle, transmission, epizootiology. salinity effects, temperature effects Since 1987, /"cr/lai.viw mw/mw.s (Mackin et al. 1930), the caus- sion of P. marinus in the phylum Apicomplexa. but suggest a ative agent of Dernio disease, has been the most important patho- recent common ancestrv' with the dinoflagellates. gen of the eastern oyster Crassostrea virgimca (Gmelin) along the Along the east coast of the United States prior to the late 1 980s. east coast of the United States south of Delaware Bay. The origin P . marinus was restricted to high salinity portions of coastal bays of P. marinus is obscure, but it probably always has been an and estuaries south of Delaware Bay, although it apparently was associate of oysters. It was first reported in Chesapeake Bay oys- absent from the seaside bays of the eastern shore of Virginia and ters in 1949 (Andrews and Hewatt 19571. The pathogen was first Maryland (Andrews 1988). In the Chesapeake Bay, P. marinus described as Dermocxstidium marinum because of apparent affin- was prevalent in the lower Bay. but was restricted to the mouths of ities with fungal parasites of freshwater fishes (Mackin et al. the major tributaries in Virginia and southern Maryland (Fig. I). 1950). It was later reclassified as Labyrinthomyxa marina because There were a few localized concentrations oi P. marinus in Mary- of observations of gliding cells similar to those present in slime land, primarily in Fishing Bay and Eastern Bay. The pathogen was molds (Mackin and Ray 1966). Ultrastructural observations (Per- observed locally in Delaware Bay in the mid-1950s, as a result of kins 1976) of an apical complex in the motile zoospore stage led importing infected oysters from Chesapeake Bay, but it never Levine (1978) to establish the new genus Perkinsus for the patho- caused significant mortality in oysters and appeared to die out as gen within the phylum Apicomplexa. Taxonomic placement of P. importations stopped in the late 1950s (Ford 1992). North of Del- marinus in the Apicomplexa has been controversial because of the aware Bay the parasite was absent or at least undetectable, presence of a number of morphological and life cycle character- In endemic areas P . marinus has always been responsible for istics more typical of the Mastigophora (tlagellates) than of the some oyster mortality, but it did not significantly affect harvest Apicomplexa (Vivier 1982). Molecular sequence data (Fong et al. most years because of the large natural sets on public beds and 1993. Goggin and Barker 1993) and a recent phylogenetic analysis good seed-oyster availability for private planters in Virginia. An based on sequence data (Siddall et al. 1995) do not support inclu- excellent review of the history of research on this pathogen and of 17 18 BURRESON AND RaGONE CaLVO Kilometers Figure I. Distribution of P. marinus in Chesapeake and Delaware Bays prior to its spread during the 1980s. E = Eastern Bay, F = Fishing Bay, M = Mobjack Bay, P = Pocomoke Sound, S = mouth of St. Mary's River, T = Tangier Sound. Data from Andrews (1981) and Krantz and Otto (1981). P. marinus epizootiology in Chesapeake Bay prior to the late 1980s was provided by Andrews ( 1988). It has long been known that the distribution and local abun- dance of P . marinus are controlled by environmental conditions (Andrews 1988). During the late 1980s and early 1990s the dis- tribution and epizootiology of P. marinus in the Chesapeake Bay deviated from historical patterns as the result of four consecutive drought years and concomitant warm winters from 1985 to 1988. During that period, P. marinus spread to all productive oyster grounds in Chesapeake Bay either by natural processes or by movement of infected oysters. Elevated salinities and warm win- ters allowed the pathogen to survive in areas that historically were disease-free. Although drought conditions have abated and rainfall patterns have returned to more or less typical conditions, with wet winters and springs, especially during 1993 and 1994. P. marinus continues to persist tenaciously in most areas of the Chesapeake Bay. The presence of the pathogen throughout the James River seed area in Virginia has been especially troublesome because infections develop to lethal levels when seed oysters are trans- planted to high salinity growout areas. Epizootiology of p. marinus in Chesapeake Bay 19 The puqjosc of this review is to summarize the current distri- bution oi P. inannus. to discuss the current understanding of the epizootiology of the pathogen in Chesapeake Bay. with emphasis on changes since 14X5, and to discuss the impact of this pathogen on the oyster resource of the Chesapeake Bay. We will focus on environmental controlling factors, as they are particularly impor- tant in Chesapeake Bay. and we will attempt to identify areas where data are especially lacking and where research needs to be focused. PRESENT DISTRIBUTION OF P. MARIMS As of late 1994. P inaiiiius is known from as far north as Wellfleet Harbor. Cape Cod Bay. MA (Ford 1996), south through- out the bays and estuaries along the east coast of the United States, including virtually all oyster beds in Delaware and Chesapeake Bay. and throughout the Gulf of Me.xico as far south as Tabasco, Mexico (Burreson et al. 1994a. Soniat 1996). In the mid-l98()s, P iminnus had not been reported north of Chesapeake Bay; thus, the present distribution represents either a major northward expansion of P. marinus or a significant increase in abundance of the parasite in areas where it may have been present but was undetectable. Although P. marinus was reported periodically trom native oysters in Delaware Bay in the mid-1950s, probably as a result of importing infected oysters from Chesapeake Bay or other southern areas, it never became established and has never been responsible for significant oyster mortality (Ford 1992). This situation changed in 1990 when P . marinus became abundant in Delaware Bay and was also found in Great Bay along the Atlantic Coast. Abundance and distribution within Delaware Bay increased during 1991 and significant oyster mortality occurred then and in subse- quent years. Prevalence of P. marinus in New Jersey coastal bays during 1991 ranged from 30% in Dry Bay. Manasquan and Tuck- erton. 50% in Raritan Bay and 85% in Great Bay. As of 1994. the parasite is abundant on all oyster beds on the north shore of Del- aware Bay. including the seed beds (Fig. 2); it seems to be much less abundant along the southern. Delaware shore (Ford 1996). In Delaware Bay, it appears that P. marinus spread from un- detected localized foci as a result of unusually warm winters dur- ing the period (Ford 1992), although effluent into the Maurice River from shucking houses processing P. /?k7(7>!i/.v-infected oys- ters from the Gulf of Mexico may have also contributed to the spread. Because of the drought conditions and concomitant warm winters, the pathogen was able to become established and it has now replaced Haplosporidnmx nelsoni (MSX). although perhaps temporarily, as the most important oyster pathogen in Delaware Bay (Ford 1996). The spread of P. marinus northward into Long Island Sound was probably also facilitated by warm winters. The pathogen may have spread from undetected localized foci of infection established in the past by importation of infected oysters from southern areas, but possibly also by recent movement of infected oysters, although recent movements have not been documented. Prevalence and in- tensity of P. marinus are high in oyster samples from some areas of Long Island Sound and the south shore of Cape Cod, for ex- ample Cotuit. MA, and oyster mortality attributed to this pathogen has been relatively high in some areas (Ford 1996). In the Chesapeake Bay. P. marinus spread into historically low salinity areas during the prolonged drought of the late 1980s and it is now present on all public oyster beds in both Virginia and Maryland (Figs. 1 and 2). although significant oyster mortality is restricted to those areas where salinity is above about 12 ppt for most of the summer and fall. The parasite is also now present in the bays along the seaside of the eastern shore of Virginia and Maryland, probably as a result of moving infected oysters to those locations from Chesapeake Bay. Unfortunately, Virginia scientists were not aware of the spread of P. marinus during 1985 and 1986. Dr. Jay Andrews had retired m 1984 and the oyster disease monitoring program that had been underway since 1959 was terminated. The first indication of the spread was very high mortality in September 1986 in oysters trans- planted from the James River seed area to three tributaries along the south shore of the Potomac River, the Coan and Yeocomico Rivers and Machodoc Creek. Disease analyses revealed high lev- els of P . marinus in all three areas (>90% prevalence, 2.6-3.4 weighted prevalence). The source of the seed was revealed as Miles ground in the lower portion of the James River seed area (Fig. 3) and subsequent analyses of oysters from that site revealed high prevalence (96%) of P. marinus although most infections were light (weighted prevalence = 1.36) (Burreson 1987). It be- came clear that the parasite had spread into the lower seed areas and had been moved to the growout areas in infected seed oysters. The drought conditions allowed P . marinus to spread into the seed area and also allowed it to flourish in the growout areas because salinity was favorable (> 12 ppt) in those areas as well. During the seven year period from 1985 through 1991. only 1989 was con- sidered a wet year. The growout tributaries off the south shore of the Potomac River had previously been free of significant mortal- ity caused by P . marinus although the parasite was observed in these areas during some years (Andrews 1981). Oysters in the lower portion of the James River seed area were known to harbor P. marinus periodically (Andrews and Hewatt 1957), but preva- lence and intensity were always low. By 1988, intensity of P. marinus infections in endemic areas had increased dramatically and oyster mortality was high, espe- cially during 1987 and 1988. In addition, favorable salinities al- lowed the pathogen to spread into new areas and by 1991 P. marinus had spread to most oyster growing areas of the Chesa- peake Bay including Maryland either by natural processes or by movement of infected oysters (Table 1). Oysters in previously non-enzootic areas were highly susceptible to P. marinus, infec- tion prevalence and intensity were unusually high, and mortality was high on both planted grounds and public beds in favorable salinity. The parasite was present at Wreck Shoal (WS) (Fig. 3) in the middle of the James River seed area in 1986 and had spread to Deepwater Shoal (DWS), the uppermost oyster bed in the James River by 1988. Prevalence and intensity of P. marinus continued to increase in the James River through 1991. Similarly, the patho- gen spread throughout the Rappahannock River and was present at Ross Rock, the uppermost oyster bed by 1992. although both prevalence and intensity were very low at that site. A similar up-bay spread of P. marinus occurred in Maryland through the 1980s and early 1990s (Figs. 1 and 2) from foci of infection in Tangier Sound, Holland Strait, Tar Bay and near the mouth of the St. Mary's River. By 1987 the parasite had spread up the main stem of the Bay to Swan Point north of the mouth of the Chester River and throughout Fishing Bay and the mouth of the Choptank River. In the Potomac River the parasite spread to the mouth of Clements Bay during 1987. During 1988 P. marinus spread throughout the Choptank and Little Choptank Rivers and further up the Potomac River lo the mouth of the Wicomico River. By 1992 the pathogen had spread throughout the Chester River and 20 BURRESON AND RaGONE CaLVO Kilometers f r ^NORFOLK ^ Figure 2. Present distribution of P. marinus in Chesapeake and Delaware Bays. Shading indicates maximum annual prevalence. E = Eastern Bay, F = Fishing Bay, M = Mobjacit Bay, P = Pocomoke Sound. S = mouth of St. Mary's River, T = Tangier Sound. Data from Ragone Calvo and Burreson (1995), G. E. Krantz (personal communication) and S. E. Ford (personal communication). was present on every productive oyster bar in Maryland (Krantz 1993). Intensity of infections during summer and fall increased each year in previously invaded areas and oyster mortality was greater than 50% in areas with favorable salinity including most areas south of Kent Point (Krantz 1990. Krantz 1992. Krantz 1993). The spread of P. marinus into areas in the lower Chesapeake Bay where it was historically absent seems to have been a long- term acquisition. Unusually high spring runoff during 1993 and 1994, a very wet July in 1994 and a cold winter in 1993-94 had little effect on the subsequent fall prevalence of P. marinus in the James River. VA (Table I), although intensity of infections de- clined somewhat from a peak in 1991 . Prevalence and intensity of P marinus infections did decline to a greater extent in the upper Bay in Maryland during 1994 (Table 1). The historical absence of P. marinus in the upper Bay and upper reaches of the major tributaries suggests that the pathogen will eventually be eliminated from these areas if normal environmental conditions of cold win- Epizootiology of p. marinus in Chesapeake Bay ^/^^^^>^^ 10 KILOMETERS Figure 3. James River, V A, showing localidiis of \ari()us moniloring stations. ters and wet springs continue, hut monthly monitoring in Virginia during the 1990s has demonstrated that the decline will be slow and may take a decade or more. Unfortunately, with the present widespread distributron of P . mariiuts. any drought period will allow the pathogen to increase in abundance and will only prolong the problem. South of Chesapeake Bay, P . marinus has always been present in bays and estuaries including intertidal oyster beds. The drought conditions of the late 1980s also caused a dramatic increase in abundance of P inwimis in North Carolina. Oyster mortality at- tributable to P mannus was first documented in the fall of 1988 in southern North Carolina. From 1988 through 1992 the pathogen TABLE 1. Prevalence (% infected) off. marinus at various locations in Chesapeake Bay" Location 1980 1986 1989 1991-92" 1994 Virginia James River, Wreck Shoal James River. Horsehead Rock James River. Deepwater Shoal Rappahannock River. Broad Creek Rappahannock River. Sniokcy Point Rappahannock River. Bowlers Rock Rappahannock River. Ross Rock Maryland Swan Point Chester River. Old Field Eastern Bay, Bugby Choptank River, Cooks Point Choplank River. Sandy Hill Patuxent River. Broomes Island Potomac River. Cornfield Harbor Potomac Rover, Ragged Point Potomac River. Lower Cedar Point Holland Straits Tangier Sound, Old Woman's Leg 0 0 100 0 0 48 0 0 8 04 84 44 0 04 44 nd 0 40 0 0 0 0 0 03 0 0 10 0 0 100 0 0 23 0 0 53 50 50 57 75 nd nd 0 nd 93 0 0 03 80 nd nd SO nd 23 100 100 88 100 100 88 24 23 37 100 100 100 100 100 90 10 100 100 100 96 56 64 46 16 0 03 20 63 90 83 40 77 10 83 57 73 ■" Data from Andrews ( 1981 ). Burreson ( 1987, 1490, 1992. 1W3), Krantz ( 1990. 1992. personal communication). Krantz and Otto ( 1981 ) and Ragone Calvo and Burreson ( 1995). nd = no data. " For most locations, either 1991 or 1992 was the year of highest prevalence. 22 BURRESON AND RaGONE CaLVO spread northward along the eastern edge of Pamlico Sound and then across to the western side, eventually infecting all oyster beds and causing high inortality. Oyster mortality from P. marinus continued during 1993 and 1994 in Pamlico Sound, but mortality seems to have declined in southern areas near Bogue Sound (M. Marshall, personal communication). The status of P . marinus in South Carolina and more southern states does not seem to have changed significantly from historical levels although there have been few data published on distribution and intensity in these areas (Burtcll et al. 1984, Crosby and Rob- erts 1990) and extensive disease monitoring is lacking. The patho- gen is present and causes some oyster mortality in most areas. ANNUAL CYCLE OF P. MARINVS PREVALENCE AND INTENSITY IN CHESAPEAKE BAY Samples of non-spat oysters from natural oyster beds exhibit a pronounced seasonal cycle in both prevalence and intensity (ex- pressed as weighted prevalence) of P . mannus infections when diagnosed with the tluid thioglycollate technique (Ray 1952. Ray 1966) of mantle, gill and rectal tissue (Fig. 4). Typically, preva- lence and weighted prevalence of P . marinus infections begin to increase in June. On average, maximum values of both parameters are reached in September; prevalence at WS (Fig. 4) reaches 1009^ every year and weighted prevalence reaches 2.0 with some years above 3.0. These values contrast with 1954 when prevalence of 60% and weighted prevalence of 1.5 were considered intense in- fections (Andrews and Hewatt 1957). The relative contribution of multiplying overwintering infections and acquisition of new infec- tions to this increase is not clear, but based on timing of P. mari- nus transmission (see below) it appears that the early summer increase is primarily the result of proliferation of overwintering infections. Prevalence may remain high through January, but in- tensity, measured as weighted prevalence, usually declines sharply in October if peak values are above 2.5. This decline is probably due, at least in part, to death of heavily infected oysters. Preva- lence and weighted prevalence values decline through the winter and reach minimum values in late spring, typically April or May (Fig. 4). Since 1988 some infections have been found throughout the winter and spring in Virginia except at locations where salinity becomes less than about 5 ppt for extended periods. Detectable overwintering infections are contrary to the situation prior to 1985 when P. marinus infections were either absent or undetectable during winter (Andrews 1988), and are probably the result of the much higher abundance of the parasite since 1985. However, the winter/spring decline in prevalence is in part an artifact of the low sensitivity of the standard fluid thioglycollate medium (FTM) technique (see Fig. 5). Recently, Ragone Calvo and Burreson (1994) using antibody detection and Bushek et al. (1994) using total body burden fluid thioglycollate analyses have also shown that prevalence does not decline as dramatically during winter as routine FTM assay would suggest. However, intensity does de- cline during late winter and spring and all infections durmg that period are of very low intensity (Fig. 5) (see also Bushek et al. 1994). The decline in intensity is partly the result of mortality of moderately and heavily infected oysters during winter, but inten- sity values decline even in areas where infection intensity is rel- ativelv low and where no mortalitv occurs. It is not known if this Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 4. Annual cycle of P. marinus prevalence (top) and infection intensity, expressed as weighted prevalence (bottom), in Chesapeake Bay oysters. Dotted lines demonstrate year-to-year variability for years 1988-94. Bold line represents the average of all years, 1988-94, Prevalence and intensity v»ere determined using the FTM method de- scribed by Ray (1966), Oysters (n = 25) v^ere sampled monthly from Wreck Shoal, James River, VA. 100 90 80 70 60 50 40 30 20 10 - 1 1 ^ Total body burden |~~| Ray tissue diagnosis u 100 90 80 70 60 I 50 91 ru 40 ^ CD 30 20 10 0 2 Oct Nov Dec Jan Feb Mar Apr May Jun Jul Figure 5. Seasonal prevalence (top) and intensity (bottom) off. mari- nus as determined by standard Ray tissue FTM assays and by total body burden estimations. Intensity is expressed as weighted preva- lence for standard FTM assays (right axis) and as loglO-transformed cells per gram wet tissue weight for body burden estimations (left axisl. Oysters were sampled from Wreck Shoal, James River, VA, Sample size was 25 for standard FTM assays and 20 for total body burden assays. Epizootiology of p. marinus in Chesapeake Bay 23 decline is the result of active defense processes by the oyster or passive processes related to tolerance of P. marinus to temperature and salinity although previous researchers have suggested that par- asite cells are actively eliminated (Andrews 1988). Maximum and minimum prevalence and weighted prevalence values during the annual cycle and the timing of increases and decreases are affected by the local temperature and salinity re- gimes (sec next section) although the general pattern remains con- sistent in all areas. For example, while prevalence values have reached 100% at Wreck Shoal in the James River every fall since 1988, they reached 100% at Horsehead Rock, an upriver station in lower salinity, only during 1991. In addition, annual prevalence values usually peak one or two months later and begin to decline one to two months earlier in these low salmity areas. There seems to be an oyster size threshold required for P. marinus infection in nature. Spat less than about 30 mm in shell height are rarely found infected using routme fTM assay (Burre- son 1991). This apparent size threshold may be the result of low sensitivity of FTM assay, but more likely is due to reduced filter- ing capacity of small oysters. It does not appear to be the result of innate resistance to the parasite, as Andrews and Hewatt (1957) have demonstrated that small oysters acquire infections when dose is high. Infections with P . marinus are known to be dose depen- dent and small oysters probably don't filter enough water to ac- quire sufficient infective stages of the parasite in nature. ANNUAL CYCLE OF P. MARINUS-VSDVCED OYSTER MORTALITY In areas of favorable salinity (>I2 ppt) oyster mortality result- ing from P. marinus infections usually begins about the first of August and continues through early winter although most oysters die in late August and September. The proliferation of F. marinus is temperature dependent and abundance within an oyster increases so long as temperature is above about 20°C; thus, an unusually warm spring or fall will prolong the development period of the pathogen and result in greater oyster mortality. The mortality pattern of oysters placed into salinity regimes conducive to parasite development depends on the prior history of P. marinus infection. Uninfected oysters larger than about 30 mm shell height usually acquire P. marinus infections during mid to late summer of the first year. Mortality is usually low because declining water temperature during fall prohibits development of most infections to lethal levels, but mortality as high as 407f may occur if oysters are about 50-60 mm shell height. High mortality, often greater than 90%, will occur in these oysters during the second summer if environmental conditions are favorable for P. marinus development (Fig. 6). This mortality pattern is drastically different than that prior to 1985 when significant oyster mortality from P. marinus did not occur until the third summer after initial infection. Management strategies proposed by Andrews and Ray (1988) to harvest oysters after two summers of growout were successful prior to the 1980s, but have not been as effective since 1986 because high mortality occurs during the second summer of exposure. Spat that are less than about 30 mm shell height during late summer/early fall will usually not acquire P. marinus that summer and they can often be grown to market size before significant mortality from P. marinus occurs. Aquaculturists can reduce mor- tality caused by P. marinus by spawning oysters late and delaying 100- 90 80 70 60 50 40 30 20 10 H. nelsoni ■ Heavy n Moderate D Light I L U L _ L U L U May Aug May Jul Sep P marinus ■ Heavy □ Moderate D Light J J A 1990 Figure 6. Disease-associated cumulative oyster mortality (bottom) and prevalence and intensity of H. nelsoni (MSX) (topi and P. marinus (middle) in hatchery-reared juvenile oysters deployed In the lower York River, VA. Prevalence of H. nelsoni was very low, especially during 1990, so mortality can be attributed to /'. marinus. For parasite data, prevalence is indicated by total bar height and percentage of sample in each intensity category by shading, .Sample size = 25, Ar- rows indicate samples examined but no infections found. Two oyster stocks are compared in each graph — upper James River (U) and lower James River (L|, Mean shell height in July 1989 was 42 mm for both groups. placing them in P. marinus-enzootic waters until late September. In this situation most spat avoid infection by P. nuirinus but still grow well until winter. They will acquire P. marinus infections during the next summer, but mortality will be low. Experience has shown that oysters can reach market size by the following spring and can be harvested before high mortality results the following summer (M. Luckenbach, Virginia Institute of Marine Science [VIMS], personal communication). Because P. marinus is present on all seed-oyster bars in the Chesapeake Bay, oysters should not be moved from seed areas to high salinity growout areas. Light infections will intensify and high mortality will almost certainly occur the first summer after transplantation. Nor should P. marinus-mfected oysters be moved to low salinity with the expectation that the pathogen will be eradicated. Monthly monitoring in the upper James River, VA (Ragone Calvo and Burreson 1994), has clearly shown that P. marinus can survive long periods (weeks to months) of salinity below 5 ppt and days to weeks in fresh water. 24 BURRESON AND RaGONE CaLVO ENVIRONMENTAL CONTROL OF P. MARINVS INFECTIONS Salinity Clearly, salinity is an important environmental control of P . mahnus because prevalence and intensity of tlie pathogen within an estuary increase with increasing salinity (Andrews 1988. Craig et al. 1989, Soniat and Gauthier 1989). Historically. P. manims was absent from Chesapeake Bay waters with summer salinities of about 15 ppt or less and a large proportion of oyster grounds located in the upper reaches of Chesapeake Bay tributaries were disease-free. As a consequence of four consecutive drought years 1985-88 the abundance and distribution of P. mahnus increased dramatically and the parasite became present on all oyster grounds in Virginia. The historical restriction oi P . mahnus to high salinity areas (>12-15 ppt) suggests that over the long term the parasite cannot tolerate the low salinities of the upper Bay or upper tribu- taries; however, since its spread in the late 1980s the parasite has persisted in most of these lower salinity areas despite the return to normal and even below normal salinities. In 1987, VIMS initiated an intensive survey program to mon- itor P . mahnus prevalence and intensity at three oyster bars in the upper James River, VA, which, prior to the drought years of 1985-88, were free of P. marinus. Since 1987. oysters (n = 25) have been sampled monthly from Wreck Shoal (WS). Horsehead Rock (HH), and Deepwatcr Shoal (DWS) (Fig. 3). These bars are located along a salinity gradient with average salinities for the years 1987-94 of 14ppt(±4.3, n = 318)at WS, 9ppt(±4,l, n = 166) at HH, and 7 ppt (±4.0, n = 245) at DWS. As a con- sequence of abnormally high salinities associated with below av- erage streamflows, P. mahnus invaded WS in the summer of 1986 and within a year prevalence at the site was 100%. The parasite spread upriver to HH during the summer of 1987 and was first observed at DWS m the summer of 1988. P. marinus spread through HH and DWS more slowly than at WS, but since 1990 peak fall prevalences have ranged from 40 to 88% at DWS and from 88 to 100% at HH (Fig. 7l. In addition to affecting the local distribution and abundance of P. marinus. salinity also has a significant effect on P. marinus infection acquisition and intensity. Paynter and Burreson (199 found that juvenile cultured oysters deployed at a low salinity site (8-10 ppt) did not acquire infections while those at moderate ( 12- 15 ppt) and high (16-20 ppt) salinity sites did acquire infections. Furthermore, infection intensity at the moderate salinity site was lower than that at the high salinity site. Similarly, the limiting effect of low salinity on P. marinus prevalence and intensity is observed in native James River oyster populations. At WS, the area having the highest salinity, infections overwinter at a higher prevalence and intensity and increase as the water temperature warms at a much faster rate than at the lower salinity areas. HH and DWS (Fig. 8). During the summer months infections in WS oysters generally progress to moderate and heavy intensity in re- sponse to high temperatures and salinities and disease-associated mortality results. For instance, during the late summer and fall months of 1994 salinity at WS ranged from 12 to 20 ppt and moderate to heavy P. marinus infections were observed in 12- 30% of the oysters sampled each month (Fig. 8). Prevalences and infection intensities decrease in an upriver direction from WS in- dicative of the limiting effect of low salinity on P. marinus. Gen- erally, only a few moderate to heavy infections are observed at HH and infections at DWS rarely progress to moderate and heavy intensity. This was apparent in 1994 (Fig. 8) when summer and fall salinities ranged from 8 to 15 ppt at HH and from 5 to 12 ppt at DWS. While environmental fluctuations may alter the severity of P. marinus epizootics from year to year, the general trend of increasing prevalence and intensity in a downriver direction per- sists. 90- DWS Salinity =7 ppt 60- 70- □ Light E 60^ g 50- ^ 40- H Moderate 1 Heavy 30- 20- 10- 0- n n t ^ * * -^ n-i 1 1 ' 1 — 1 1 1 1 HH Salinity = 9 ppt Figure 7. Prevalence of P. marinus in oysters sampled along a salinity gradient in the upper James River, VA, Oysters (n = 25) were sampled monthly from \VS, HH and DWS. Diagnoses were made using stan- dard FTM assays. Average salinities for period 1987-94 al VV'S, HH and DWS were, respectively, 14, 9 and 7 ppt. Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 8, P. marinus prevalence (total bar height) and percentage of sample in each intensity category (shading) in oysters sampled along a salinity gradient in the upper James River. VA. in 1994. Oysters (n = 251 were sampled monthly from WS. HH and DWS. Diagnoses were made using standard Ray tissue FTM assays and infection intensity was categorized as light, moderate and heavy. Arrows indicate sam- ples examined but no infections found. Salinity is the average for 1994 based on two to three observations per months. EpIZOOTIOLOGY of p. MARINUS IN CHESAPEAKE BaY 25 Based on these studios o{ P manniis intection patterns along the James River salinity gradient and in other Chesapeake Bay tributaries, critical salinity regimes for P . mannus activity can be defined. These studies indicate that: 1 ) if summer and fall salinities are consistently less than 9 ppt. P . marinus may persist but infec- tions are limited to light intensity and no oyster mortality results: 2) if summer and fall salinities vary from 9 to 15 ppt. some infections may progress to moderate and heavy intensity, but as- sociated oyster mortality is relatively low: and 3 1 if summer and fall salinities are consistently greater than 15 ppt. moderate to heavy infections may be numerous and oyster mortality may be relatively high. Preliminary statistical analysis of the relationship of salinity and P. mannus infection intensity and prevalence in James River oysters was conducted using a Spearman rank correlation analysis. The analysis was based on 1 80 observations which included monthly determinations of prevalence for a five year period. 1990-94. at three oyster beds. W.S. HH. and DWS (Fig. }). Twenty-five oysters were collected from each site each month and examined for P. luannus by culturing rectal, gill, and mantle tissue in FTM following the method described by Ray (1966). Infection intensities were ranked as light, moderate, and heavy and assigned numerical values of 1 . 3 and 5 according to the scale of Mackin (1962). The numerical intensity values, which included 0 for negative diagnoses, were then averaged for the determination of weighted prevalence. Salinity was recorded at each site 1-3 times each month and monthly means were determined. The results of the correlation analysis demonstrated a highly significant (p < 0.0001) and strong correlation between salinity and P mannus prevalence and intensity in James River oysters (Spearman rank corrected rho = 0.729 and 0.727. respectively) (Fig. 9). A subsequent linear regression analysis indicated that salinity accounts for 51% of the variability in prevalence. Only a limited number of field studies employing statistical analyses of data have been conducted. Significant positive correlations be- tween salinity and P . marinus prevalence and intensity have been observed in Gulf Coast oysters (Soniat 1985. Craig et al. 1989. Soniat and Gauthier 1989) and in South Carolina oysters (Crosby and Roberts 1990). In the Gulf of Mexico, salinity (0-34 ppt) was observed to account for only 20% of the site-to-site variability in P. marinus infection (Craig et al. 1989) and in South Carolina salinity (29 to 35 ppt) was only weakly correlated with infection (Kendall rank tau = 0.094) (Crosby and Roberts 1990). The present analysis of the relationship between salinity and P. mari- nus activity in the James River suggests that salinity may play a more significant role in regulating P. mannus in the Chesapeake Bay than in more southern waters, although differences in the correlation results may also be attributed to differences in salinity regime and in experimental design, particularly sampling fre- quency. More rigorous statistical analysis of James River data should help to further our understanding of the role of salinity in regulating P. mannus prevalence and intensity. The association of salinity with P. marinus prevalence and intensity has been addressed by several researchers. Mackin (1951) suggested that high flushing rates, typical in the upper reaches of estuaries, dilute infective pathogen cells thereby limit- ing the ability of water-borne infective stages to infect oysters. Thus, the absence of P. marinus from low salinity areas was attributed to the absence or scarcity of infective cells (Ray and Mackin 1954. Mackin 1956, Andrews and Hewatt 1957). An- drews ( 1988) related the Chesapeake Bay distribution of P. mari- Jan Apr Jul Oc! Jan Apr Jul Ocl Jan Apr Jul Ocl Jan Apr Jul Ocl Jan Apr Jul Oct 1990 1991 1992 1993 1994 Figure 9. P. marinus prevalence (solid line) and mean monthly salinity (dotted line! at DWS. HH and VVS. James River, VA. Oysters (n = 25) were sampled monthly and prevalence was determined using standard Ray tissue FTM a.ssays. Mean monthly salinity was calculated from measurements recorded one to three times each month. nus to the physical circulation dynamics of specific Chesapeake Bay tributaries. He asserted that large flushing-type rivers, such as the James River, are not as favorable to the pathogen as small coastal plain tributaries, such as the Choptank River, because large discharges of fresh water during the winter and spring reduce the period of salinities favorable to P marinus and dilute the concen- tration of infective cells. While dilution of infective particles by fresh water discharge may be an important factor influencing the distribution off. mari- nus It is not entirely responsible for the reduced disease levels observed in low salinity areas. In recent years many laboratory investigations have documented an inhibitory effect of low salinity on various aspects of P. marinus epizootiology. Scott et al. (1985) found lower mortality in infected oysters held in the laboratory at 8-10 ppt than in oysters held at 21-25 ppt. Their results support the work of Ray ( 1954) in which P. marinus in artificially infected oysters tolerated low salinity ( 1 0-1 3. 5 ppt) exposure but develop- ment of infections and subsequent mortalities of oysters were de- layed relative to the high salinity (26-28 ppt) control group. Ragone and Burreson (1993) exposed naturally infected oysters from the upper James River. VA. to three low salinity treatments. 6. 9 and 12 ppt. and found no reduction in P. mannus prevalence at any of the treatments after eight weeks of exposure. However, development of infection was retarded at 12 ppt compared to the high salinity (20 ppt) control and infection intensity did not in- crease at 6 and 9 ppt. Oyster mortalities after eight weeks were 26 BURRESON AND RaGONE CaLVO significantly less at 6 and 9 ppt than at 12 or 20 ppt. which were nearly equivalent. This study suggests that 9-12 ppt is a critical range for P. marimis activity supporting recent field observations. Although P. marinus infection progression may be limited by low salinity. P. marinus is quite tolerant of low salinities, unlike H nelsoni which is intolerant of salinities less than 10 ppt (Ford 1985). Chu et al. (1993) succeeded in artificially establishing in- fections in oysters maintained in the laboratory at salinities as low as 3 ppt. Prevalences of P. tnarinus five weeks after challenge by mantle cavity injection with lO'' meronts were 50. 70 and 82% at 3. 10 and 20 ppt, respectively. All infections observed at 3 ppt and most found at 10 ppt were of low intensity, suggesting that parasite proliferation within the host was limited relative to the high salin- ity control. Differences in oyster mortality and infection progression be- tween high and low salinity environments may be attributed to the direct effect of salinity on host and/or parasite physiology. Several in vitro investigations have helped us gain a better understanding of the direct effect of salinity on P . marinus. Perkins ( 1966) and Chu and Greene ( 1989) found that low salinity inhibited sporula- tion of prezoosporangia isolated from oyster tissue cultured in FTM. The recent, successful culture of P. marinus (Gauthier and Vasta 1993, Kleinschuster and Swink 1993. La Peyre et al. 1993) has allowed a more rigorous examination of the salinity tolerance of P. marinus in the absence of host influences. In vitro. P. mari- nus is tolerant of a wide range of salinities and has been reported to proliferate at osmolalities from 340 to 1930 mOsm ( 10-60 ppti (Dungan and Hamilton 1995). Osmolalities below 340 mOsm were not tested. Ma.ximal proliferation was observed at 790 mOsm (25 ppt) and near-ma.\imal proliferation occurred within the range of 475-960 mOsm (15-30 ppt) (Dungan and Hamilton 1995). While cultured P. marinus cells exhibit growth at salinities as low as 10 ppt they are relatively intolerant of acute hypoosmotic shock. Burreson et al. (1994b) exposed P. marinus meronts harvested from 22 ppt culture media to 0. 3. 6, 9, 12 and 20 ppt artificial seawater. After a 24 hour exposure period at 28°C. viability was assessed using the vital stain neutral red. Percent mortality was 99% at 0 ppt. 90% at 3 ppt. 70% at 6 ppt. 43% at 9 ppt. 20% at 12 ppt and <5% at the 22 ppt control treatment. The effect of salinity on percent mortality was highly significant. When the osmotic concentration of the various seawater treatments was ad- justed with sucrose to the equivalent of 22 ppt. percent mortality was low in all treatments and no different than the optimal control condition demonstrating that low salinity-induced mortality was caused by a decrease in osmotic pressure, not a decrease in sodium or another important ion. The low survival of cultured P. marinus cells at 6 ppt is surprising considering the documented ability of the parasite to survive in oysters at 6 ppt and 20°C for a period of eight weeks (Ragone and Burreson 1993) and may not be relevant to natural conditions in which changes in osmotic condition are likely to be more gradual and mediated by host responses. In summary, within the Chesapeake Bay region P. marinus activity is greatly influenced by salinity. Prevalence and intensity of the pathogen intensify during drought years during which low stream flows cause above average salinities in upper tributary wa- ters. In general, prevalence and intensity of P. marinus increase in a downriver direction. Infections are restricted to low intensity in areas consistently having salinities of less than 9 ppt. while high intensity infections and associated oyster mortality often occur during the summer and fall in areas with salinities greater than 12-15 ppt. Once established in a low salinity area the parasite tenaciously persists and has been observed to tolerate salinities <5 ppt for a period of at least three months and to quickly respond to exposure to favorable salinities as evidenced by increases in prev- alence and intensity. Laboratory studies have demonstrated that P. marinus survival, infection progression and pathogenicity are sa- hnity limited, supporting recent field observations. Temperature Temperature appears to be the most important environmental factor affecting the large scale geographic distribution of P. mari- nus (Ray and Mackin 1954. Andrews and Hewatt 1957. Quick and Mackin 1971 ). The northern limit of P. marinus is believed to be controlled by minimum winter temperature (Andrews 1988). Max- imum summer temperatures and/or the duration of temperatures above 20-25°C are probably also important, but the role of min- imum or maximum temperature on the geographic distribution has not been rigorously investigated. Within the Chesapeake Bay, seasonal temperature changes are largely responsible for the seasonal periodicity of the annual P. marinus cycle. Winter temperatures, which on average (1947-90) are below 5°C for eight weeks, are associated with a regression in tissue infection levels resulting in spring minimums in infection intensity and prevalence. Infections begin to intensify in late spring as water temperature exceeds about 20''C and parasite pro- liferation occurs (Andrews 1988). In Chesapeake Bay. tempera- tures favorable to parasite proliferation. >20°C, occur for about 20 weeks and temperature may exceed 25°C for a period of 10 weeks. The highest parasite prevalences and intensities are ob- served in September immediately following maximal summer tem- peratures. The occurrence of high prevalences and intensities at high temperature most likely reflects temperature associated in- creases in parasite multiplication rate but may also relate to tem- perature associated depressions in host defense capabilities and physiological condition. In high salinity environments, infections intensify to lethal levels and mortality usually occurs from August through October Infection intensity declines as temperatures de- cline in winter (Figs. 10 and 11). however, the parasite is known to persist patently at temperatures as low as ()-5°C (Andrews 1988). Winter water temperature in the Bay typically averages 4-5°C. but may be as low as 1°C or less for extended periods during unusually cold winters. Body burden analysis allowed the documentation of a remarkable decline from December to May in number of meronts per gram wet weight of oyster tissue; however, prevalence remained at 90-100% (Fig. 5). These residual infec- tions rapidly proliferate as temperatures rise in late spring. Numerous field and laboratory experiments have focused on the relationship between temperature and P. marinus infection intensity and prevalence. P. marinus infection in South Carolina oysters was significantly but weakly correlated with temperature (Kendall rank correlation coefficient = 0.283) (Crosby and Rob- erts 1990); temperature explained 16.7% of the variability in in- fection intensity. This result contrasts with those of Burrell et al. ( 1984) and Craig et al. ( 1989) in which no statistically significant relationship between temperature and intensity was found in South Carolina and Gulf of Mexico oysters, respectively. Differences in these results have been attributed to differences in frequency and interpolation of temperature measures (Crosby and Roberts 1990). In the James River tributary of the Chesapeake Bay. P. mari- nus intensity and prevalence clearly follow seasonal fluctuations in water temperature (Fig. 10). Preliminary statistical examination of Epizootiology of p. marinvs in Chesapeake Bay 27 Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oc! Jan Apr Jul Oct 1990 1991 1992 1993 1994 Figure 10. P. marinus prevalence (solid line) and mean monthly water temperature (dotted line) at DWS. HH and WS. James River, VA. Oysters (n = 25) were sampled montlily and prevalence was deter- mined using standard FTM assays. Mean monthly temperature was calculated from temperature measurements recorded at six minute intervals at the VIMS York River monitoring station. this relationship was conducted using a Spearman rant; correlation analysis. Data for P . marinus are the same as described for the statistical analyses of salinity relationships. Mean monthly water temperature was calculated from water temperatures recorded at six minute intervals by a continuous metering system at the VIMS York River monitoring station. The initial correlation analysis failed to find a statistically sig- nificant relationship between mean monthly water temperature and P . marinus prevalence or intensity (p > 0.05). This result agrees with the findings of Soniat (1985) and Craig et al. (1989) and contrasts with the findings of Crosby and Roberts (1990) which indicated a statistically significant but weak correlation between water temperature and P. marinus intensity. However, when the James River data set was reanalyzed with temperature lagged by two to four months, a significant correlation between water tem- perature and P. marinus prevalence and intensity was found. The relationship was strongest when the temperature was lagged three months (eg. April prevalence and January temperature): water temperature was strongly and significantly correlated with both prevalence (Spearman rho = 0.704, p < 0.001) and weighted prevalence (Spearman rho = 0.706, p < 0.001). Regression anal- ysis indicated that when lagged three months temperature ex- plained 39% of the variability in prevalence and 46% of the vari- ability in weighted prevalence. The contribution of temperature to the year-to-year variability in P marinus activity is not well understood. Minimum winter temperature, while thought to control the geographic distribution of the pathogen, is not clearly associated with year-to-year vari- ability of P. marinus epizootics in the Chesapeake Bay. During an eight year ( 1987-94) monthly parasite survey of James River oys- ters (described above, see salinity discussion) extreme above and below average fluctuations in winter temperature were observed. The relationship of these temperature fluctuations to the subse- quent summer epizootics is somewhat obscure, in part because it is difficult to separate the effects of salinity fluctuations from temperature fluctuations. The coldest winters in terms of average winter temperature and duration of weekly average temperatures below 5°C were the winters of 1987-88 and 1993-94 (Fig. 11). Regardless of the cold winter, subsequent summer prevalences and intensities in 1988 were among the highest recorded, but this was also an abnormally dry year. In 1994 winter water temperatures were below 5°C for a period of eight weeks and 1-2°C below the long term average for six of the eight weeks. This unusually cold temperature regime seemed to have little negative impact on P. marinus as 1994 had the third highest average summer intensity and prevalence was still greater than 96% at WS for a five month period from August through December (Fig. 1 1 ). During the win- ter of 1989-90 record low temperatures were observed in Decem- ber but after the first week of January water temperatures were generally above average. The low December temperatures may have contributed to the relatively early decline of overwintering infections and to the relatively slow rise in prevalence during the summer; however, salinity was also relatively low during the pe- riod. Abnormally warm winter temperatures may have a more sig- nificant impact on P. marinus activity than abnormally cold tem- peratures. The winter of 1991 was the warmest winter during our survey period (Fig. 11). Mean weekly temperature never went below 5-6°C during the winter and temperatures were 1-3°C above the long term average throughout the year. Overwintering prevalences were low. but prevalence rapidly increased with the onset of warm summer temperatures and remained above 90% for six months at WS (Fig. 11). making 1991 the worst year in terms of average summer prevalence. The cool fall of 1990 and low 1990 prevalences may be responsible for the low 1990-91 overwinter- ing levels, while the abnormally wann winter water temperatures combined with dry conditions probably caused high summer prev- alences in 1991. However. 1991 summer temperatures were also above average and probably also contributed to the high summer parasite level (Fig. 11). The association between temperature and P. marinus infection has been the focus of several laboratory investigations. Andrews and Hewatt (1957) reported that at 15°C the development of es- tablished infections was retarded and new infections did not ap- pear. Similarly Fisher et al. (1992) found infection progression and oyster mortality were reduced in oysters held at 18°C com- pared to those held at 27°C. More recently, Chu and La Peyre ( 1993) exposed oysters held at 10, 15, 20 and 25°C to P. marinus through mantle cavity injections of lO*" meronts obtained from infected oyster tissue. Infections were observed in oysters from all treatment groups; however, prevalence declined with decreasing temperature and moderate and heavy infections were only ob- served in oysters at 20 and 25°C. Forty-six days after challenge P. marinus prevalence was 23% at IO°C, 46% at 15°C, 91% at 20°C and 100% at 25°C. The influence of temperature on P. marinus infection intensity and prevalence may relate to host and/or parasite activity. Both 28 BURRESON AND RaGONE CaLVO 1988 1990 Figure II. Year-to-year variability of water temperature (A), James River streamflov* (B), and prevalence (P) and weighted prevalence (\VP) of P. marinus (C). In Panel A, ^^eekly temperature averages for each year are contrasted with the long term mean weekly temperature for the period 1947-94. In Panel B, monthly streamflow averages for each year are contrasted with the long term mean monthly streamflow for the period 1951-95. Water temperature is from the continuous record of the \ IMS York River Monitoring Program; streamflow data were obtained from the U.S. Geological Survey. cellular and humoral oyster defense activities have been shown to be affected by environmental temperature (Fisher 1988, Chu and La Peyre 1989. Chu and La Peyre 1993). Unfortunately, the role of these putative oyster defense activities in combating P . marinus remains speculative. In vitro culture of P. marinus has afforded an opportunity for analysis of temperature effects on the parasite in the absence of host intluences. Proliferation of P. marinus in culture was near maximal at room temperatures between 15 and 35°C and optimal at SST. Minimal proliferation occurred at 10 and 40°C, and no proliferation occurred at 4°C (Dungan and Ham- ilton 1995). In a study conducted by Burreson et al. ( 1994b I cul- tured P. marinus cells were quite tolerant of a 24 hour exposure to temperatures as low as TC at high salinity (22 ppt). Survival at 1 and 5°C treatments was greater than 90% based on a vital stain assay and did not significantly differ from that at 10. 15 and 28°C treatments (Burreson et al. 1994b). Temperature appeared to have a greater effect when exposed cells were in lower salinity condi- tions. A general trend of higher mortality at lower temperature was observed at 6. 9 and 12 ppt, however, the temperature effect was only significant at 9 ppt. It was suggested that cold temperatures may inhibit metabolic processes such as free amino acid release which may enable some cells to survive at lower salinities. Prior to the culture of P. marinus. in vitro studies on the effect of temper- ature were limited to the assessment of sporulation of presporangia isolated from thioglycollate-incubated infected tissue. Viability was determined by the presence of motile spores within sporangia. Sporulation was optimal at 25-35°C (22 ppt) (Perkins 19661. Both the maximum percent sporulation and the rate of sporulation were greatly reduced at 20°C (60% of optimal) and no sporulation oc- curred at temperatures less than I8°C (Perkins 1966). In a similar experiment Chu and Greene ( 1989) observed that prezoosporangia survived at 4°C for up to four days but did not survive below 0°C for one day. In summary, it appears that in the Chesapeake Bay region P. marinus activity and annual periodicity are largely controlled by seasonal temperature fluctuations. This conclusion is supported by a strong statistically significant correlation between temperature and P. marinus prevalence and intensity. It is difficult to precisely define the effect of temperature on year-to-year variability of P. marinus infections based on analyses conducted to date. However some trends are apparent. Abnormally cold winter temperatures may hasten the decline in infection intensity during the winter EPIZOOTIOLOGY of p. MARINVS IN CHESAPEAKE BaY 29 months and delay the rise in prevalence during the summer months, but they have little impact in reducing the severity of summer epizootics. Conversely, abnormally warm winter temper- atures probably increase the severity of summer epizootics. The interaction of temperature and salinity is probably more important than either factor acting alone. Recent evidence suggests that, in Chesapeake Bay, the prevalence and intensity oi P. mari- iiiis decline much more rapidly during winter in low salinity areas (18 ppt) (Ragone Calvo and Burreson 1994). Laboratory investigations are needed to deter- mine the synergistic effect of temperature and salinity fluctuations on the progression and/or regression of established P. mariiuis infections. TR.4NSMISSION DYNAMICS Although it is well documented that transmission of P. manmis is direct from oyster to oyster and that any life cycle stage of P. mcuinus seems capable of initiatmg infections in the laboratory (Ray 1954, Andrews 1988), the natural dynamics of transmission are poorly understood. Transmission is dose dependent and it seems to take unusually high concentrations of any life cycle stage to initiate infections (Andrews 1988). Transmission is thought to occur through the digestive tract because initial foci of infection occur in the gut epithelium (Mackin 1951). although this obser- vation needs confirmation with careful laboratory studies. In any case, the cell type that actually initiates infection and the mecha- nism of infection are poorly understood. The role of flagellated, free-swimming zoospores in initiating infections in nature is es- pecially problematic. Zoospores certainly don't seem to be re- quired for initiation of infections, as infections result from expo- sure to isolated meronts or even minced, infected oyster tissue. However, the transformation that may occur after merozoites or minced tissue are added to an aquarium or injected into the mantle cavity of an oyster are unknown. The occurrence of early infec- tions in the stomach epithelium suggests ingestion of infective stages, not penetration of gill or mantle by zoospores. But perhaps zoosporulation occurs in the gut lumen and released zoospores penetrate in localized areas of the gut epithelium. It appears that a very high dosage of zoospores, on the order of I x 10', is required to initiate an infection (Andrews 1988). This dose seems high for an efficient parasite but may be an artifact of the experimental designs employed. Zoospores must have some function or their production wouldn't have evolved. Maybe they are a dispersal mechanism and are only produced in nature under certain condi- tions that are not presently understood. Inoculation of non-zoospore stages into the mantle cavity of oysters has demonstrated that meronts produce higher prevalences and higher intensities of infection than prezoosporangia (Volety and Chu 1994). but the pattern and process of infection were not followed in these studies. A high proportion of P. mannus cells in an oyster occur within host hemocytes and it has been proposed that hemocytes that scavenge the epithelial surface of the oyster gut lumen phagocytose ingested P. mariims cells and then mi- grated through the epithelial layer and into the oyster carrying the parasite with them. An innovative study with intubated fluorescent polystyrene beads has demonstrated that such events do occur (Alvarez et al. 1992). Once inside an oyster and under favorable environmental conditions, P. marinus multiplies within hemo- cytes, eventually killing the hemocyte and releasing the P. mari- nus cells. These cells are phagocytosed by other hemocytes and the cycle repeats; eventually pathogen cells are carried throughout the oyster. The developmental cycle of P. marinus within oysters is relatively well understood and recently has been reviewed by Perkins ( 1991. 1993), but studies that examine the initial infection process in oysters are critically needed. Even though it is known that transmission is direct, in nature there is a poor understanding of the source of infective stages, the dose required to initiate infections and the duration of the infection window. The prevailing conceptual model is that transmission occurs during periods of high oyster mortality in summer and early fall as infective P. marinus cells are disseminated upon death and decomposition of infected oysters (Andrews 1988). However, dead, gaping oysters are consumed rapidly by scavengers (Hoese 1964) and probably don't decompose naturally and release P. marinus cells into the water. The parasite can survive passage through the gut of scavengers (Hoese 1964), but the role of scav- engers in spreading infections is unclear. In the Gulf of Mexico, transmission of P. marinus can occur via the ectoparasitic snail Boonea impresso (White et al. 1987), but for the Chesapeake Bay no vectors have been identified. Dissemination of P. marinus in fecal matter from live oysters seems likely, given the destruction of gut epithelium observed in live, heavily infected oysters, but is poorly documented. Mackin (1962) proposed that P. mannus overwinters as a free hypnospore in the sediment and that annual epizootics are initiated by release of infective cells in the spring. Andrews (1988) countered that if this were true, imported unin- fected oysters should develop infections in June or early July rather than late July or August as he observed. Nevertheless, the presence of a saprobic stage should not be ruled out. Recently, flow cytometric techniques have been developed that may allow quantification of disseminated P. marinus cells in the water col- umn (Roberson et al. 1993). Such data should provide insight into seasonality of infection pressure. Field experiments to assess the timing of infections are under- way at VIMS. Separate groups of uninfected oysters are being exposed in the lower York River for two week periods throughout the year and are then warmed in the laboratory for four weeks to allow infections to develop to detectable levels. Results for 1994 indicate that the highest infection pressure occurs during the last two weeks of August and the first two weeks of September (Fig. 12), a period that corresponds closely with maximum oyster mor- tality; however, some infections were acquired as early as late June. »50 * *...,.* n. ....•' i 1 / "n" \ \ "■■■:. *□,.,*„,* Oct Dec Apr May Jun Jul Aug Sep Figure 12, P. marinus infection acquisition (bars) in uninfected sen- tinel oysters deployed for two week periods in the lower York River, VA, and percent mortality per day of local infected oysters (line). Bars represent the prevalence of infection in sentinel oysters as determined by total body burden assays. Arrows indicate no new infections during the period. 30 BURRESON AND RaGONE CaLVO There has not been much laboratory research conducted on the effect of environmental variables on transmission dynamics, but meaningful experiments are difficult to perform m the laboratory because of the difficulty of simulating natural conditions and the artificial nature of the challenge used in most experiments. In experiments where P. marinus meronts were injected into the mantle cavity of oysters held at various temperatures and salini- ties, transmission did occur at temperatures as low as 10°C (sa- linity = 17.5 ppt) and salinities as low as 3 ppt (temperature = 21.0°C) (Chu and La Peyre 1993, Chu et al. 1993). These results clearly show that infection by P. marinus is possible at low tem- perature and low salinity conditions. However, there is little evi- dence that infections occur under these conditions in nature, prob- ably because of an absence of infective cells in the water or low oyster filtration rates. Andrews and Hewatt (1957) found that new infections were not acquired in the field at salinities ranging from 1 to 13 ppt and Paynter and Burreson ( 1991) found no infection acquisition in the field at salinities ranging from 8 to 12 ppt. These results suggest that although it is possible for P. marinus to infect oysters at relatively low temperature and salinity conditions, such transmission probably does not occur in nature. However, it must be remembered that diagnosis during these studies was by routine FTM assay; more sensitive diagnostic techniques may yield dif- ferent conclusions. Movement of P. marinu.s into historically low salinity areas occurred during drought periods when salinities were elevated. Now that the pathogen is present and persisting on all oyster beds it is important to determine if transmission is occurring in areas where salinity has returned to more normal conditions. Field stud- ies using uninfected sentinel oysters in low salinity areas are prob- ably the best method to determine whether transmission is occur- ring in low salinity areas. Other critical research areas for trans- mission dynamics include elucidation of the early infection process including cell type and infection site, timing of infections in nature and the role of environmental variables. THE ROLE OF OYSTER DEFENSE MECHANISMS Considering all of the research that has been conducted on the role of oyster defense mechanisms in controlling P. marinus in- fections it is perhaps surprising how little is known about the topic. If oyster defense capabilities do control levels of P. marinus in oysters, the components and mechanisms involved have not been identified so it is impossible to assess the role of defense mecha- nisms in the epizootiology of P. marinus disease. Unfortunately. most studies have been correlative studies where some putative defense mechanism such as lysozyme or agglutinin is measured and correlated with intensity off. marinus infections. These stud- ies have produced much useful data on components of the defense mechanisms in oysters and their relation to environmental param- eters, but they have not demonstrated any direct link between pathogen levels and serum or cellular components, perhaps be- cause of the high variability of measured parameters and because correlation analysis, even when it is statistically significant, doesn't necessarily demonstrate cause and effect. Because of typ- ical highly variable results, investigators have been reluctant to rule out any component, but recently Chintala et al. ( 1994) have demonstrated, and clearly stated, that the particular serum agglu- tinins studied play no role in defense against P. marinus. More innovative, directed studies are needed to determine the role of the various hemolymph components that have been postu- lated as important in defense against P. marinus. Experimental manipulation of the defense components and subsequent monitor- ing of infection progression, compared to untreated control oys- ters, should shed light on the role of specific hemolymph compo- nents. For example, employing monoclonal antibodies as blocking agents of putative defense components, passive transfer of purified components into oysters or incubation of P. marinus with serum components prior to injection into oysters may enable determina- tion of the roles of these components. Unfortunately, much pre- liminary research may have to be done before meaningful studies can be conducted. One critical need is to determine how P. marinus avoids intra- cellular killing by hemocytes. Although P. marinus is not an ob- ligate intracellular parasite, most cells in oysters are found within hemocytes. Obviously, hemocytes recognize the parasite as for- eign and phagocytose individual cells. However, there seems to be no intracellular killing of the parasite or at least the multiplication ability of the parasite during summer far outweighs any killing. Rather, the parasite multiplies within hemocytes and eventually bursts the cell membrane releasing more individual cells that get phagocytosed by other hemocytes and carried throughout the oys- ter in hemolymph. Oyster hemocytes are known to produce the typical free oxygen radicals (ROIs) involved in intracellular killing by vertebrate phagocytes (Anderson et al. 1992, Anderson 1994) but they seem to be ineffective against P. marinus. at least during periods of active parasite multiplication, possibly because P. mari- nus suppresses ROl production (Volety and Chu 1995). THE ROLE OF POLLUTION AND WATER QUALITY IN P. MARINUS ABUNDANCE One of the questions most often asked of oyster disease re- searchers concerns the role of declining water quality in the in- crease of oyster diseases during the recent past. Most oystermen and many of the lay public consider pollution to be a critical factor in disease processes and blame it solely for the increase in oyster diseases. However, there is little evidence to support their claim. In the Chesapeake Bay there is no correlation between water qual- ity or the level of pollution and disease abundance. Abundance of P. marinus is just as high in relatively unpolluted areas as in polluted areas of equivalent salinity. For example, Tangier Sound is one of the areas in Maryland hardest hit by oyster diseases and it was characterized by the EPA Chesapeake Bay Study as having the best water quality in Maryland. Similarly in Virginia the abun- dance of oyster pathogens is high wherever salinity is favorable and many of the areas where oysters were decimated by disease, such as Pocomoke Sound, are relatively pristine. Pathogen abundance clearly correlates with salinity levels and the dramatic increase in abundance in the late 1980s can be ex- plained by drought conditions and resulting increased salinity with concomitant warm winters as a secondary factor. As discussed in earlier portions of this review, a variety of laboratory studies and field observations support the primary role of salinity and temper- ature in modulating P. marinus abundance. Nonetheless, even relatively unpolluted areas today are not as pristine as they were even 20 years ago so some subtle effects of pollution or water quality cannot be completely ruled out. Pollu- tion effects are known to modulate host defense mechanisms in aquatic vertebrates (Anderson 1990), but since the role of oyster defense mechanisms, if any, in controlling P. marinus infections is not understood, it cannot be concluded that pollution suppresses the oyster's ability to inhibit the pathogen. Epizootiology of p. marinvs in Chesapeake Bay 31 Although pollution clearly is not one ot the primary' factors responsible for recent increases of oyster diseases, there has been very little research on the potential subtle effects of toxicants on P . marinus disease progression. Only recently have studies been completed thai suggest some effect of toxicants on P. nuirinus disease development. Winstead and Couch (1988) reported rapid proliferation of P. imirinKs in oysters exposed to high concentra- tions (600 nigP') of the carcinogen n-nitrosodiethylamine when compared with unexposed control oysters. Chu and Hale (1994) found elevated P. marinus prevalence in oysters exposed to water soluble fractions derived from estuarine sediments grossly con- taminated with polycyclic aromatic hydrocarbons and then chal- lenged w ith P . imirimis. These results suggest some effect of these chemicals on either pathogen multiplication or host condition or defense mechanisms, but it is difficult to determine the environ- mental relevance of these studies because it is unclear how the concentrations utilized in the experiments compare to actual levels of these compounds found in the water column in nature. Tnbutyltin (TBT) has also been shown to enhance P. marinus disease progression and increase cumulative oyster mortality dur- ing experiments using environmentally relevant levels of TBT. Maximum prevalence and intensity levels of P. marinus occurred sooner in oysters exposed to !()() pptr TBT for five months and exposed to P. marinus after one month than in oysters not exposed to TBT and infected with P. marinus similarly (Anderson et al. 1995). In addition, mortality was higher in oysters that were both exposed to TBT and also infected with P marinus than in oysters either infected but unexposed or exposed but uninfected. Similar results were obtained in experiments using 30 and 80 pptr TBT (Fisher et al. 1995) although the experimental design was some- what different from that of Anderson et al. (1995) These studies suggest that environmental toxicants may have some ettect on disease development in highly polluted areas, but as the authors emphasize, there is no evidence that the dramatic increase in abundance of P marinus since 1985 is the result of increased environmental pollution. Undoubtedly, further research will better clarity the subtle interactions among oysters, disease agents and environmental contaminants. OTHER FACTORS INFLUENCING THE EPIZOOTIOLOGY OF P. MARINUS There are other factors that potentially may influence the epi- zootiology of P. marinus in the Chesapeake Bay, but little re- search has been done that is specific to the Bay. For example, recent modelling studies in the Gulf of Mexico (Powell et al. 1996) suggest that timing of food availability is important to enable oysters to outgrow the parasite (Soniat 1996). Because of the longer growing season in the Gulf of Mexico than in the Chesa- peake Bay. nutrition may be more critical in the Gulf of Mexico than in the Bay. Another factor that potentially may intluence the epizootiology of P . marinus disease in Chesapeake Bay is the well-documented summer hypoxia, but no research has been done on this interaction. It is well known that many other species of molluscs in Ches- apeake Bay harbor cells of Perkinsus sp. (Andrews 1954). The taxonomic status of Perkinsus in these other hosts has not been clarified, but if they are P. marinus then these other molluscs could serve as reservoir hosts for the pathogen. The significance of putative reservoir hosts in the epizootiology of P. marinus disease is unknown. Studies are needed to determine if lethal P. marinus infections can be induced in oysters using Perkinsus cells isolated from other mollusc species. EFFECT OF P. MARINUS ON THE OYSTER RESOURCE OF CHESAPEAKE BAY Prior to 1985 P marinus had little significant impact on the Maryland oyster industry because the pathogen was uncommon in Maryland. There were localized foci of infected oysters in the St. Mary's River in the 1960s and high mortality had decimated the local population by the late 1970s. The parasite was reported in Fishing Bay in the 1970s and in Eastern Bay in 1981 . In Virginia, where P. marinus historically was restricted to the lower Bay and mouths of major tributaries, significant, but tolerable, oyster mor- tality occurred in these areas, especially in high salinity areas or during dry years. Nonetheless, harvest in Maryland and Virginia varied between 2 and 3.5 million bushels annually during the 19,Ws. 1940s and 1950s (Fig. 13). In Maryland the harvest was primarily from public oyster beds, but in Virginia over 80% of the harvest came from private planters who planted disease-free seed oysters from the Upper James River to growout grounds in the lower Bay areas of Mobjack Bay and Hampton Roads. Productive public beds occurred in the York River and Rappahannock River, Pocomoc Sound and various small tributaries along the western shore of the Bay (Andrews 1988). Typically, private planters in Virginia held seed oysters for three years on growout grounds, but as knowledge of P. marinus epizootiology increased and it was learned that most mortality occurred during the third year, planters began harvesting after only two years of growout (Andrews 1988). This early-harvest disease-avoidance strategy worked well during the late 1950s and annual harvest in Virginia varied from 3 to 4 million bushels during that period. The sudden epizootic of H. neisoni (MSX) in Mobjack Bay beginning in 1959 (Haskin and Andrews 1988) and the resulting high oyster mortality caused private planters to eventually abandon the traditional growout areas in the lower Bay by the mid-1960s and move operations to lower salinity areas in the Rappahannock River and small tributaries along the south shore of the Potomac River such as the Coan and Yeocomico Rivers. The 1970s were generally wet and levels of both H. neisoni and P. marinus were reduced; from about 1967 through 1981 oyster harvest in Virginia was more or less stable at about 1 million bushels annually (Fig. 13) — greatly reduced from pre- 1960 levels because of reduced growout acreage resulting from abandonment of the traditional growout areas in the lower Bay. In Maryland, good spat sets coupled with low pathogen abundance because of reduced salini- 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 Figure 13. Annual market oyster landings in Chesapeake Bay from 1930 t(i 1994. 32 BURRESON AND RaGONE CaLVO ties during the 1970s produced harvests between 2 and 3 million bushels annually during the 1970s (Fig. 13). Both 1980 and 1981 were dry years; P. mahmis intensified and H . nelsoni spread into Maryland for only the second time since it first appeared in the Bay in 1959. Mortality was high in areas where salinity was favorable for the pathogens and oyster landings declined in both states from 1982 through 1984. The four consecutive drought years from 1985 through 1988 were catastrophic for oyster resources in both Maryland and Vir- ginia. As outlined above. P . marinus spread to most oyster grow- ing areas of the Chesapeake Bay including Maryland either by natural processes or by movement of infected oysters. Oysters were highly susceptible to P. marinus and mortality was high on planted grounds in Virginia and on public beds in both Virginia and Maryland. H. nelsoni also intensified during 1987 and 1988 and contributed substantially to the mortality in both states. The end result in Virginia was the virtual elimination of oysters from public beds in the lower Bay and from all but the uppermost reaches of the major tributaries. Current estimates are that less than 5% of traditional public oyster beds in Virginia are productive and these are all in the upper James River. Public beds were depleted by disease, and private planters, fearing losses from planting infected seed oysters, were (and still are) reluctant to transplant oysters for growout. Because of the absence of oysters in other areas of the lower Bay. harvesting pressure mcreased significantly in the upper James River. VA. beginning in 1986. Annual harvest increased in Virginia during 1986 and 1987 (Fig. 13) because of th large numbers of oysters in the upper James River, but the stocks were rapidly fished out and harvest plum- meted in subsequent years. Since 1988, over 95% of the public market-oyster harvest in Virginia has come from the James River. Although there have been some restrictions placed on harvesting in Virginia, no quotas have been adhered to and remaining critical broodstocks in Virginia are being fished heavily. In Maryland, approximately 79% of the harvest durmg 1993-94 came from areas north of the Bay Bridge and 66% of the harvest came from the Chester River. As a comparison, during the 1973-74 season in Maryland, only 2. 1% of the harvest came from the Chester River. These harvest figures from low salinity areas in both Virginia and Maryland demonstrate the impact of disease on the oyster resource in high salinity areas of the Chesapeake Bay. The oyster resource rebounded in Maryland during 1994—95 and harvest increased for the first time in three years. Wet springs during 1993-94 resulted in lower salinity in Maryland, reduced levels of P. marinus and increased oyster survival in high salinity. In Virginia, no improvement was observed during 1994-95. EPIZOOTIOLOGY GENERALIZED MODEL A generalized model of P. marinus epizootiology in Chesa- peake Bay is shown in Fig. 14. The model, based on data from the last decade, represents average timing of events that vary annually depending on temperature and salinity regimes. Infection regres- sion begms in November and continues through May when min- imum prevalence and intensity values are reached. Minimum in- fection parameters are reached approximately three months after minimum winter temperature and about one month after minimum salinity, although timing of minimum salinity varies much more than minimum temperature. Infection regression seems to be caused by the direct effects of temperature and salinity on P. marinus survival. The role of oyster defense mechanisms in in- fection regression is unknown but cannot be ruled out. inteclion regra,. Figure 14. Generalized summary of P. marinus epizootiology in Ches- apeake Bay. Dashed lines represent months of reduced activity com- pared to months represented by solid lines. Infection prevalence and intensity begin to increase in June as water temperature increases above 20°C and overwintering infec- tions begin to proliferate. Increase in prevalence and intensity from June through most of August seems to be due almost entirely to the proliferation of overwintering infections. Infection prolifer- ation probably continues until October in oysters that don't die from the disease. Maximum prevalence and intensity occur in September, approximately six weeks after maximum water tem- perature, if salinity is greater than about 15 ppt. but values may peak one or two months later in lower salinity. Maximum values reached depend on the salinity regime. A warm spring allows early proliferation of overwintering in- fections and oyster mortality may begin by eariy July, but under typical conditions most oysters die in August and September. Some mortality may continue at a reduced level until January or even later depending on fall temperatures. Total oyster mortality depends on the temperature and salinity regime and on the infec- tion history. New P. marinus infections are acquired by oysters shortly after oyster mortality commences and the correlation between new in- fections and oyster mortality suggests that the dying oysters are the source of infective stages. Most new infections are acquired during the last two weeks of August and the first two weeks of Septem- ber, corresponding with the period of greatest oyster mortality. Because temperatures are greater than 25°C during this period and infections develop rapidly, there may be one or more cycles of oyster mortality/new infections/proliferation between late July and early October. ACKNOWLEDGMENTS The following individuals provided unpublished data or other information on historical and present distribution of P. marinus: Susan Ford, Rutgers University; Steve Jordan and George Krantz, Epizootiology of p. makinus in Chesapeake Bay 33 Maryland Department of Natural Resources; and Mike Marshall, North Carolina Division of Marine Fisheries. Chris Judy. Mary- land Department of Natural Resources, and Jim Wesson. Virginia Marine Resources Commission, provided oyster harvest data for Maryland and Virginia, respectively. Juanita Walker provided par- asite diagnoses for the VIMS oyster disease monitoring program. Kenny Walker. Gustavo Calvo. Caroline O'Farrell. Brenda Sandy Flores. Sandra Blake and Anna Schotthoefer assisted with sample collection and processing. Gary Anderson provided York River, VA, temperature data. Collection of original data on seasonal body burden of P. nuihims was funded by USEPA under order number 3G0834NTSE; collection of data on new infection acqui- sition was funded by the NOAA Oyster Disease Research Program under grant number NA47FL0158. VIMS contribution number 1968. LITERATURE CITED Alvarez, M. R., F. E. Friedl. C, M. Hudson & R. L. O'Neill. 1992. Uptake and tissue distribution of abiotic particles from the alimentary tract of the American oyster: a simulation of intracellular parasitism. J. Imertehr. Pathol 59:290-294. Anderson, D. P. 1990. Immunological indicators: effects of environmental stress on immune protection and disease outbreaks. Anwr. Fish. Sm . Symp. 8:38-50. Anderson, R. S. 1994. Hcmocyte-denved reactive oxygen intemiedlate production in four bivalve molluscs. Dev. Comp Immunol- 18(21:89- 96. Anderson. R. S.. K. T. Payntcr & E. M. Burreson. 1992. Increased re- active oxygen intermediate production by hemocytes withdrawn from Crassostrea virginica infected with Pcikinsm muriiiii^- Biol. Bull. 183:476-481. Anderson. R. S., M. A. Unger&E. M. Burreson. 1996. Enhancement of Perkinsus marimis disease progression in TBT-exposed oysters {Cras- sostrea virginica). Mar. Environ. Res. (In press). Andrews, J. D. 1954. Notes on fungus parasites of bivalve molluscs in Chesapeake Bay. Proc. Natl. Sliellfish. Assoc. 45:157-163. Andrews. J. D. 1981. Perkinsus mannus = Dermocyslidium marinum ("Demio") in Virginia. 1950-1980. Data Report No. 16. Virginia Institute of Manne Science. Gloucester Point. Virginia. Andrews, J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marimis and its effect on the oyster industry. Amer. Fish. Soc. Spec. Piibl. 18:47-63. Andrews. J. D. & W. G. Hewatt. 1957. Oyster mortality studies in Vir- ginia II. The fungus disease caused by Dermocystulium marinum in oysters of Chesapeake Bay. Ecol. Monogr. 27:1-26. Andrews. J. D. & S. M. Ray. 1988. Management strategies to control the disease caused by Perkinsus mannus. Amer. Fish. Soc. Spec. Puhl. 18:257-264. Burrell, V. G., Jr.. M. Y. Bobo & J. J. Manzi. 1984. A companson of seasonal incidence and intensity of Perkinsus marinus between subtidal and intertidal oyster populations in South Carolina. J. World Maricult. Soc. 14:301-309. Burreson. E. M. 1987. Prevalence of the major oyster diseases in Virginia waters — 1986. A summary of the annual monitoring program. Marine Resource Report 87-1. Virginia Institute of Marine Science. Gloucester Point. Virginia. Burreson. E. M. 1990. Status of the major oyster diseases in Virginia — 1989. Marine Resource Report 90-1 . Virginia Institute of Marine Sci- ence. Gloucester Point. Virginia. Burreson, E. M. 1991. Effects of Perkinsus marinus infection in the east- em oyster, Crassostrea virginica: I. Susceptibility of native and MSX- resistant stocks. J. Shellfish Res. 1 0{ 2 ):4 17-423. Burreson, E. M. 1992. Status of the major oyster diseases in Virginia — 1991. Marine Resource Report 92-1 . Virginia Institute of Marine Sci- ence. Gloucester Point, Virginia. Burreson, E. M. 1993. Status of major oyster disea.ses in Virginia — 1992. Marine Resource Report 93-5. Virginia Institute of Marine Science. Gloucester Point. Virginia. Burreson, E. M.. R. S. Alvarez. V. V. Martinez & L. A. Macedo. 1994a. Perkinsus marinus (Apicomplexa) as a potential source of oys- ter Crassostrea virginica mortality in coastal lagoons of Tabasco. Mex- ico. Dis. Aquat. Org. 20:77-82. Burreson. E. M.. L. M. Ragone. J. F. La Peyre. F. Counts & K, T. Paynter Jr. iy94b. Acute osmotic tolerance of cultured cells of the oyster pathogen Perkinsus marinus (Apicomplexa: Perkinsida). Comp. Biochem. Physiol. 109A(3):575-582. Bushek. D.. S. E. Ford & S. K. Allen. Jr. 1994. Evaluation of methods using Ray's fluid thioglycollate medium for diagnosis of Perkinsus marinus infection in the eastern oyster. Crassostrea virginica. Ann, Rev. Fish Dis. 4:201-217. Chmtala. M. M.. S. E. Ford. ^W. S. Fisher & K. A. Ashton-Alcox. 1994. Oyster serum agglutinins and resistance to protozoan parasites. J. Shellfish Res. 13(1): 1 15-121. Chu. F,-L. E. & K. H. Greene. 1989. Effect of temperature and salinity on in vitro culture of the oyster pathogen. Perkinsus mannus (Api- complexa: Perkinseal. J. Invertehr. Pathol. 53:260-268. Chu, F.-L. E. & R. C. Hale. 1994. Relationship between pollution and susceptibility to infectious disease in the eastern oyster, Crassostrea virginica. Mar. Environ. Res. 38:243-256. Chu. F.-L. E. & J. F. La Peyre. 1989. Effect of environmental factors and parasitism on hemolymph lysozyme and protein of Amencan oysters {Crassostrea virginica). J. Invertehr. Pathol. 54:224-232. Chu. F.-L. E. & J. F. La Peyre. 1993. Perkinsus marinus susceptibility and defense-related activities in eastern oysters Crassostrea virginica: temperature effects. Dis. Aquat. Org. 16:223-234. Chu. F.-L. E.. J. F. La Peyre & C. S. Burreson. 1993. Perkinsus marinus infection and potential defense-related activities in eastern oysters, Crassostrea virginica: salinity effects. J. Invertehr. Pathol. 62:226- 232. Craig, A., E. N. Powell. R. R.Fay&J. M.Brooks. 1989. Distribution of Perkinsus marinus in Gulf Coast oyster populations. Estuaries 1 2:82-91 . Crosby. M. P. & C. F. Roberts. 1990. Seasonal infection intensity cycle of the parasite Perkinsus marinus (and an absence of Haplosporidium spp.) in oysters from a South Carolina salt marsh. Dis. Aquat. Org. 9:149-155. Dungan, C. F. & R. M. Hamilton. 1995. Use of a tetrazolium-based cell proliferation assay to measure effects of in vitro conditions on Perk- insus marinus (Apicomplexa) proliferation. J. Eukanolic Microbiol. 42(4):379-388. Fisher. W. S. 1988. Environmental influence on bivalve hemocyte func- tion. Amer. Fish. Soc. Spec. Puhl. 18:225-237. Fisher. W. S.. J. D. Gauthier & J. T. Winstead. 1992. Infection intensity of Perkinsus marinus disease in Crassostrea virginica (Gmelin. I79I) from Gulf of Mexico maintained under different laboratory conditions. J. Shellfi.sh Res. 1 1(2);363-369. Fisher. W. S.. L. M. Oliver. E. B. Sutton. C. S. Manning & W. W. Walker. 1995. Exposure of eastem oysters to tributyltin increases the severity of Perkinsus marinus disease. J. Shellfish Res. 14:265-266. Fong. D.. R. Rodriguez. K. Koo, J. Sun, M. L. Sogin. D. Bushek. D. T. J. Littlewood & S. Ford. 1993. Small subunit ribosomal RNA gene sequence of the oyster parasite Perkinsus marinus. Mol. Mar. Biol. Biotechnol. 2(6):346-350. Ford. S. E. 1985. Chronic infections of Haplosporidium nelsoni (MSX) in the oyster Crassostrea virginica. J. Invertehr. Pathol. 45:94—107. Ford, S. E. 1992. Avoiding the spread of disease in commercial culture of molluscs, with special reference to Perkinsus marinus (Demio) and Haplosporidium nelsoni. J. Shellfish Res. 11:539-546. Ford. S. E. 1996. Range extension by the oyster parasite Perkinsus mari- 34 BURRESON AND RaGONE CaLVO nus into the northeastern United States; response to chniate change? J . Shellfish Res. 15:45-46. Gauthier, J. D. & G. R. Vasta. 1993. Continuous in vitro culture of the eastern oyster parasite Perkinsiis mariinis. J. Imertebr. Pathol. 62: 321-323. Goggin. C. L. & S. C. Barker. 1993. Phylogenetic position of the genus Perkinsus (Protista: Apicomplexa) based on small subunil rihosomal RNA. MoI. Biochem. Parasitol. 60:65-70. Haskin, H. H. & J. D. Andrews. 1988. Uncertainties and speculations about the life cycle of the eastern oyster pathogen Haplosporidiiim nelsoni (MSX). Amer. Fish. Sac. Spec. Publ. 18:5-22. Hoese, H. D. 1964. Studies on oyster scavengers and their relation to the fungus. Dermocystidiwn mariimm. Proc. Natl. Shellfish, .\ssoc. 53: 161-174. Kleinschuster. S. J. & S. L. Swink. 1993. A simple method for the in vitro culture oi Perkinsus marinus. The Nautilus 107(21:76-78. Krantz, G. E. 1990. Maryland oyster population status report. 1989 fall survey. CBRM-OX-90-1 . Maryland Department of Natural Resources. Oxford, Maryland. Krantz, G. E. 1992. Maryland oyster population status report. 1991 fall survey. CBRM-OX-92-1. Maryland Department of Natural Resources. Oxford, Maryland. Krantz, G. E. 1993. Maryland oyster population status report. 1992 fall survey. CBRM-OX-93-3. Maryland Department of Natural Resources. Oxford, Maryland. Krantz, G. E. & S. V. Otto. 1991. More on oyster diseases, p. 68-83. In: Webster, D. (Ed. I, Oyster Culture in Maryland 1980. A Proceedings. College Park, University of Maryland Sea Grant. La Peyre, J. F., M. Faisal & E. M. Burreson. 1993. In vitro propagation of the protozoan Perkinsus nwrinus. a pathogen of the eastern oyster, Crassostrea virginica. J. Eukaryotic Microbiol. 40:304-310. Levine, N. D. 1978. Perkinsus gen.n. and other new taxa in the protozoan Phylum Apicomplexa. J. Parasitol. 64(3):549. Mackin, J. G. 1951. Histopathology of infection of Crassostrea virginica (Gmelin) by Dermocystidiwn marinum Mackin, Owen and Collier. Bull. Mar. Sci. Gulf Caribbean 1:72-87. Mackin, J. G. 1956. Dermocystidium marinum and salinity Proc. Natl. Shellfish. Assoc. 46:116-128. Mackin, J. G. 1962. Oyster disease caused by Dermocystidium marinum and other microorganisms in Louisiana. Publ. Inst. Mar. Sci. Univ. Te.xas 1:132-229. Mackin, J. G., H. M. Owen & A. Collier. 1950. Preliminary note on the occurrence of a new protistan parasite, Dermocystidium marinum n sp. in Crassostrea virginica (Gmelin). Science 111:328-329. Mackin, J. G. & S. M. Ray. 1966. The taxonomic relationships of Dcr- mocystidium marinum Mackin, Owen and Collier. J. Invertebr. Pathol. 8:544-545. Paynter, K. T. & E. M. Burreson. 1991. Effects of Perkinsus muriiiii.s infection in the eastern oyster. Crassostrea virginica: II, Disease de- velopment and impact on growth rate at different salinities. J Shellfish Res. I0(2):425-43l. Perkins, F. 0. 1966. Life history studies of Dermocystidium marinum. an oyster pathogen. Ph.D. Dissertation, Florida State University. Talla- hassee, Florida. Perkins, F. O. 1976. Zoospores of the oyster pathogen, Dermocystidium marinum. I. Fine structure of the conoid and other sporozoan-like organelles. J. Parasitol. 62:959-974. Perkins, F. O. 1991. "Sporozoa": Apicomplexa, Microsporidia, Haplos- poridia, Paramyxea, Myxosporidia, and Actinosporidia, p. 261-331. In: Harrison, F. W. & J. O. Corliss (Eds.), Microscopic Anatomy of Invertebrates. New York, Wiley-Liss. Perkins, F. O. 1993. Infectious diseases of molluscs, p. 255-287. In: Couch, J. A. & J. W. Foumie (Eds.), Pathobiology of Marine and Estuarine Organisms. Boca Raton, CRC Press. Powell, E. N., J. M. Klinck & E. E. Hofmann. 1996. Modeling diseased oyster populations II. Triggering mechanisms for Perkinsus marinus epizootics. J Shellfish Res. 15:141-165. Quick. J. A. & J. G. Mackin. 1971, Oyster parasitism by Z,afcvriHr/iomv.ra marina in Florida. Fla. Dep. Nat. Resour. Prof. Paper Ser. 13:1-55. Ragone. L. M. & E. M. Burreson. 1993. Effect of salinity on infection progression and pathogenicity of Perkinsus marinus in the eastern oys- ter. Crassostrea virginica (Gmelin). J. Shellfish Res. 12(1): 1-7. Ragone Calvo, L. M. & E. M. Burreson. 1994. Characterization of over- wintering infections of Perkinsus marinus (Apicomplexa) in Chesa- peake Bay oysters. J. Shellfish Res. 13(1): 123- 1 30. Ragone Calvo. L. & E. M. Burreson. 1995. Status of the major oyster diseases in Virginia — 1994. Marine Resource Report 95-5. Virginia Institute of Manne Science. Gloucester Point, Virginia. Ray, S. M. 1952. A culture technique for diagnosis of infections with Dermocystidium marinum Mackin, Owen and Collier in oysters. Sci- ence 116:360-361. Ray, S. M. 1954. Biological studies of Dermocystidium marinum. a fun- gus parasite of oysters. Rice Institute Pamphlet Special Issue, Novem- ber 114 pp, Ray, S. M. 1966. A review of the culture methods for detecting Dermo- cystidium marinum. with suggested modifications and precautions. Proc. Natl. Shellfish. As.soc. 54:55-59. Ray, S. M. & J. G. Mackin. 1954. Studies on transmission and pathoge- nicity of Dermocystidium marinum. 1. Texas A&M Research Founda- tion Project 23 Technical Report. I-I I Roberson, B. S., T. Li & C. F. Dungan. 1993. Flow cytometric enumer- ation and isolation of immunofluorescent Perkinsus marinus cells from estuarine waters. J. Shellfish Res. 12:138. Scott, G. I . D. P. Middaugh & T, I. Sammons. 1985. Interactions of chlorine-produce oxidants (CPO) and salinity in affecting lethal and sub-lethal effects m the eastern or American oyster, Crassostrea vir- ginica (Gmelin), infected with the protistan parasite, Perkinsus mari- nus. p. 351-376. In: Vemberg, F. J., F. P. Thurberg, A. Calabrese & W. B. Vemberg (Eds.), Marine Pollution and Physiology: Recent Ad- vances. University of South Carolina Press. Siddall. M. E.. N. A, Stokes & E. M. Bun-eson. 1995. Molecular phy- logenetic evidence that the phylum Haplosporidia has an alveolate ancestry. MoI. Biochem. Evol. I2(4):573-58I . Soniat. T. M. 1985. Changes in levels of infection of oysters by Perkinsus marinus. with special reference to the interaction of temperature and salinity upon parasitism. Northwest Gulf Sci. 7(2):I71-I74. Soniat. T. M. 1966. Epizootiology of PfrA:mi«imarmH5 disease of eastern oysters in the Gulf of Mexico. J. Shellfish Res. 15:35^3. Soniat, T M. & J D. Gauthier. 1989. The prevalence and intensity of Perkinsus marinus from the mid northern Gulf of Mexico, with com- ments on the relationship of the oyster parasite to temperature and sahnity. Tulane Stud. Zool. Bot. 27(l):21-27. Vivier, E. 1982. Reflexions et suggestions a propos de la systematique des sporozoaires: Creation d'une classe des Hematozoa. Protistologica 18: 449-457. Volety, A. K. & F.-L. E. Chu, 1994. Comparison of infectivity and pathogenicity of meront (Trophozoite) and prezoosporangiae stages of the oyster pathogen Perkinsus marinus in eastern oysters, Crassostrea virginica (Gmelin, 1791). J. Shellfish Res. 13(2):521-527. Volety, A. K. & F.-L. C. Chu. 1995. Suppression of chemiluminescence of eastern oyster (Crassostrea virginica) hemocytes by the protozoan parasile Perkinsus marinus. Dev. Comp. Immunol. I9(2):135-142. White, M. E., E. N. Powell, S. M. Ray & E. A. Wilson. 1987. Host-to- host transmission of Perkinsus marinus in oyster {Crassostrea virgin- ica) populations by the ecloparasitic snail Boonea impressa (Pyra- midellidae). J. Shellfish Res. 6(1): 1-5. Winstead. J. T. & J. A. Couch. 1988. Enhancement of protozoan patho- gen Perkinsus marinus infections in Amencan oysters Crassostrea vir- ginica exposed to the chemical carcinogen n-nitrosodiethylamine (DENA). Disc. Aquat. Org. 5:205-213. Joiunul of Shellfish Rcsi'unh. Vol. 15, No I. 35-43, 19%, EPIZOOTIOLOGY OF PERKINSUS MARINUS DISEASE OF EASTERN OYSTERS IN THE GULF OF MEXICO THOMAS M. SONIAT Department of Biological Sciences Nicholls Slate University Thibodaiix, Louisiana 70310 ABSTRACT Pcrkiiisiis manniis iMackin, Owen and Collier) Is a major cause of mortality of eastern oysters, Cnissoslicu virginica (Gmelin). along the Gulf of Mexico. The parasite is discontinuously distributed in estuaries from Tabasco, Mexico, to the Everglades of Florida. Its distribution is essentially coincident with its oyster host, although oysters survive well at salinities slightly lower than those tolerated by the parasite. Besides a low-sahnity refuge. Gulf oysters apparently adapt to parasitic challenge with high recruitment and fast growth, which allows them to respond to natural estuarine variability more quickly than the parasite can. Temperature and salinity, but particularly their interaction, are important environmental inOuences on levels of PerkinsKs in oysters. Most of the variation in levels of infection, however, are not explained by these factors. Other environmental and biological tactors must affect the levels of parasitism observed in the field. These likely include, but are not limited to, pollution and other human influences, host nutrition and growth, spawning and reproduction, age and resistance, oyster density and distribution, and disease vectors. KEY WORDS: Pcikimu.s nun inns, Cnissasuea virginica. epizootiology. Gulf of Mexico INTRODUCTION Epizootiology is that branch of biology that deals with the nature, ecology, and causes of outbreaks of animal diseases. Its pailicuiar focus therefore is on the sum of factors, including their interactions, which control the occurrence, distribution, and inten- sity of an animal disease or pathogen. Its sister science is epide- miology, which treats similar aspects of human diseases — of which we. not surprisingly, have much greater knowledge. Even within the field of epizootiology. our knowledge of diseases of aquatic organisms generally lags far behind that of insects and domesticated species. Among aquatic pathogens, however, Perk- insKs mannus (Mackin. Owen and Collier), which parasitizes the eastern oyster Crassostrea virginica (Gmclin), is one of the most intensively studied and best known. P. marinus was first described by Mackin, Owen, and Collier ( 1950) as Dermocvslidium nniriniini. because of its apparent sim- ilarity to the freshwater parasitic fungus Dermocysndnim salmanis (Davis). Later observations suggested that D. inarinum gave rise to gliding cells on "mucoid tracks" similar to slime molds and was reclassified as Lahyrinlhonnxd manna (Mackin and Ray 1966). Although ultrastructural studies by Perkins ( 1969) revealed a likeness of the parasite to the fungi, no cytoplasmic extensions or rhizoids were observed. Levine (1978) renamed the parasite P. marinus on the basis of electron microscope work by Perkins (1976). which revealed the presence of an apical complex in a motile zoospore stage. The organism now resides in the protozoan phylum Apiconiplexa. which it shares with the sporozoans (see Perkins 1996). The discovery of P. ( = Dermocxsticliiim) marinus is the legacy of Texas A&M University Project 9. a multi-institutional effort funded by a consortium of oil companies, which investigated the causes of oyster mortalities in Louisiana oil fields (Mackin and Hopkins 1958. Ray 1996. S, H. Hopkins, unpublished observa- tion). Extensive studies were conducted on the effects of crude oil. bleedwater. natural gas. drilling mud, and seismographic surveys on oysters. The general conclusion of these studies was that none of these pollutants or activities could explain the widespread oyster mortalities observed, but that high mortalities were caused by a parasite associated with high temperature and high salinity (Mackin and Hopkins 1958, 1962; see also Ray 1996). Numerous studies (e.g., Ray et al. 1953. Mackin 1956. Quick and Mackin 1971, Ogle and Flurry 1980, Soniat 1985. Craig et al, 1989) have documented that P. marinus. commonly called "Dermo." is most prevalent during the warm months in high-salinity areas and is widely distributed along the Gulf Coast. (See Table I for a chro- nology of selected studies, ) The purpose of this paper is to review the 45 years of progress regarding our knowledge of the epizooti- ology of P. marinus in the Gulf of Mexico. DISTRIBUTION ON THE GULF COAST P. marinus is discontinuously distributed in estuaries of the Gulf of Mexico from Tabasco. Mexico (Burreson et al. 1994) to the Everglades of Florida (Craig et al. 1989; Fig. 1). Its distribu- tion in the Gulf of Mexico is essentially coincident with that of its oyster host, although the host grows well at salinities slightly lower than that tolerated by the parasite (Mackin 1956). It is un- likely that any mesohaline Gulf estuary with even modest oyster populations is free of the parasite; indeed, the long-term absence of the parasite from an oyster population is more noteworthy than its presence. A recent southern limit of the parasite has been established by Burreson et al. ( 1994). who found it in the Carmen, Machona. and Mecoacan Lagoons of Tabasco. Mexico, Prevalence (or percent infection, PI) and disease intensity (or weighted incidence, WI) ranged from 609( and 0.6 al Rio San Felipe (Carmen Lagoon. 15 ppt) to 100% infection and a Wl of 3.1 at Los Jimenez (Machona Lagoon. 32 ppt). (Reports on prevalence in oyster populations tend to be underestimates due to the possibility of false negatives using the standard thiogycollate test.) Mackin (1962) reported "Dermo" from Tampico Bay. Mex- ico, which until recently was the southernmost published record for the parasite, Perkinsus has not been extensively sampled from the lagoons north of Tampico Bay to the Rio Grande estuary near Port Isabel. Texas. Hildebrand (personal communication), how- ever, indicates that C. E. Dawson found lOO'/r infections and "high intensities" of the parasite in oyster populations from La- guna Tamiahau and that the parasite is likely found throughout the oyster-producing lagoons of the region. 35 36 SONIAT TABLE I. A chronology of selected epizootiological and related studies on P. marinus from the Gulf of Mexico. Reference Comments Mackin et al 1950 Mackin 1951 Ray 1952 Mackin 1953 Ray 1953 Ray 1954a Ray 1954b Dawson 1955 Mackin 1956 Mackin & Bos well 1956 Mackin 1962 Mackin & Sparks 1962 Ray 1966a Ray 1966b Perkins & Menzel 1966 Quick & Mackin 1971 Beckert el al. 1972 Hofstetter 1977 Ogle & Flurry 1980 Soniat 1985 White et al. 1987 Soniat & Gaulhier 1989 Soniat el al. 1989 Craig et al. 1989 Gauthieret al 1990 Wilson et al. 1990 Gauthier & Fisher 1990 Powell et al 1992 Burreson et al, 1994 Powell et al 1994 Baraiaria Bay. LA Baralana Bay. LA Pensacola. FL Gulf port, MS Aransas Bay, TX LA. FL Barataria Bay. LA Grand Isle. LA LA. TX LA. TX Apalachicola Bay. FL Baratana & Terrebonne BaMns. LA Tampico, Mexico Aransas Bay, TX Galveston Bay. TX Barataria & Terrebonne Basins. LA Biloxi. MS Mobile Bay. AL Pensacola. FL Cedar Key. FL Lower Barataria Bay. LA Freeport. TX Galveston Bay. TX Terrebonne Basin. LA Barataria Basin. LA E. of Miss, River. LA Apalachicola Bay. FL Santa Rosa Sound. FL Tampa Bay. FL TX. LA. FL Galveston Bay. TX Apalachicola Bay. FL FL AL Galveston Bay. TX Mississippi Sound. Bilovi Bay. Horn Island. MS Galveston Bay, TX Lake Borgne. LA Galveston Bay. Port Aransas. TX Galveston Bay. TX Lake Calcasieu. Vermillion Bay, Terrebonne Bay. Barataria Bay. Adams Bay. E of Miss. River. Lake Borgne. LA Biloxi Bay, MS Galveston Bay, TX Gulf of Mexico Lake Calcasieu. Terrebonne Ba- sin. Baratana Basin. E of Miss, River, LA Gulf of Mexico W. Galveston Bay, S. Padre Island. TX; Lake Borgne, LA Gulf of Mexico Tabasco. Mexico Galveston Bay, TX Mackin. Owen and Collier descnbe D mannum from histological sections of Gulf oysters; type locality is Sugar House Bend in southern Baratana Bay, The histological consequences of infection corroborates "Dermo"" as a major cause of oyster mortality. Ray develops a culture technique which greatly aids epizootiological studies. Heavy infections of "Dermo" are found m 86*^ of gapers and only l^< of live controls. Young oysters are shown to be less susceptible to "Dermo" than older ones. "Dermo" is transmitted to uninfected oysters by proximity and feeding. Culture, transmission, pathogenicity, distribution, epizootiological. and host-specificity studies are re- ported More than 50^f of oysters from 19 stations infected. 1 \^c show heavy mfections: "Dermo" is wide- spread in the Bay. The salinity tolerance of "Dermo" is nearly as great as the host, freshwater dilution of infective elements and higher concentrations of dissolved materials are hypothesized as important. A detailed life cycle is presented "Dermo" is reponed from Tampico. Mexico — no records southward. "Dermo" is not found in Copano Bay, Lower Laguna Madre. Matagorda Bay. and Atchafalaya Bay. No information available south of Cedar Ke\. FL Mortalities of oysters in trays are associated with "Dermo" intensity. Itninfected samples found only at Smith's Point (Galveston Bay) and St. Vincent'* Reef (Apalachicola An apparent absence of "Dermo" prompts a modification of the Ray (1952) technique. A motile planont (zoospore) stage is described. Planonts are apparently infective. Eighty-six stations are sampled The parasite was found "concomitantly with the host throughout the year and at salinities from 6 ppt to 36 ppt." "Dermo" is widespread in Mobile Bay, Sampling of 17 of the major reefs shows "Dermo" to be widespread. Four reefs sampled monthly for 2 years show typical late summer maxima and wmter minima at nonepi- zoolic levels, One reef (April Fool) is sampled monthly for 2 years. Wl is more closely conelated with the mteraction of temperature (T) and salinity (S) than T or S alone. The snail B. impressa transmits "Dermo" to uninfected LA oysters. May to January sampling shows lion Bav (Southwest Pass). I correlation of S. but not T. with Wl "Dermo" not found in Vermil- May and June sampling of 16 reefs shows "Dermo" absent from some reefs in Trinity and East Bays. "Dermo" is found at all 49 sites, from South Bay (near Brownsville, TX) to Faka Union Bay (Ever- glades, FL); three regional foci of infection are identiHed; north central TX. central LA. and S.W. FL. "Dermo" is sampled in four watersheds and is absent at only one low-S site; Wl is correlated with S, digestive gland atrophy, and the occurrence oi Nemaiopsis spp "Dermo" distribution is linked to land use and pollution patterns. A quantitative hemolymph assay is developed A 4-year Gulf-wide study shows that Wl is strongly correlated with long-range weather patterns. "Dermo" is found in Carmen. Machona, and Mecoacan lagoons, thus establishing Tabasco, Mexico, as the southern limit. An energy-llow model produces realistic simulations of disease progression and greatly aids understand- ing of the relationships among disease, oyster condition, and environmental factors. Pl-:is Bay, AB = Apalachicola Bay, SG = St. (Jcorge .Sound, CK = Cedar Key, TB = Tampa Bay, CH = Charlotte Harbor, NB = Naples Bay, RB = Rookery Bay, FU = Faka I'nion Bay, and EV = Everglades. See text for discussion. The parasite is present in the Port Isabel area (South Bay. lower Laguna Madre). where Craig el al. ( 1989) reported prevalences of 827f and light infections (Wi = 0.3.^). Oysters of the lower La- guna Madre are separated by a distance of 400 km from the nearest C. virgmica populations in Redfish Bay. near Corpus Christi — due to the hypersalinc conditions of the Laguna Madre (Groue and Lester 1982). The Corpus Christi Bay area was recently sampled by Craig et al. (1989). Winter samples from Nueces Bay (northern Corpus Christi Bay) had prevalences of 78% and were lightly infected (WI = 0.33), whereas winter samples from Ingleside Cove (southern Corpus Christi Bay) had a PI of 9W< and a WI of 0.33. The presence of the parasite at the northern and southern extremes suggests that it is distributed throughout the Bay. Some disagreement exists in the literature as to the occurrence oi Perkinsiis in the Aransas Bay area. For example, Hocse ( 1963) speculated on factors which might explain the absence of ■'Dermo'" from the region, yet later surveys (e.g.. Craig et al. 1989) confirm its presence. Mackin (1962) reported that the few samples taken from Co- pano Bay and Matagorda Bay were negative. Perkinsus is. how- ever, present in both Bays (Ray 1966a. Craig et al. 1989); in fact. Craig et al. ( 1989) suggest that the Matagorda Bay-Galveston Bay area is one apparent regional focus of infection in the Gulf of Mexico. Sufficient sampling with typically positive results has accumulated which suggests that Perkinsus is likely continuously distributed from Corpus Chnsti Bay to Matagorda Bay. The distribution of Perkinsus in the Galveston Bay complex (West Bay. Trinity Bay. East Bay. Galveston Bay proper) is well known (Mackin 1962. Ray 1966a. Hofstetter 1977. Soniat and Gauthier 1989. Soniat et al. 1989. Craig et al. 1989). Although occasional uninfected samples are found there (e.g.. Ray 1966a, Soniat et al. 1989). no reefs, not even those in the low-salinity waters of Trinity Bay. are consistently disease-free. Perkinsus is present in Sabine Lake, which borders Texas and Louisiana. The estuary has not been intensively sampled, and thus, the extent of its occurrence is not known. Sabine Lake is a relatively polluted, low-salinity estuary where the parasite is found at unexpectedly high prevalences, based on prevailing tempera- tures and salinities (Powell, personal communication). Craig et al. (I9S91. for example, found 1009; prevalence (WI = I.O) in a sample from lower Sabine Lake (Blue Buck Point) where the salinity was 8 ppt and when the temperature was I8°C. Gauthier et al. ( 1990) reported relatively intense infections (WT = 1.93-3.03) and high prevalences (93-1007f) at three stations along a north-south transect in Lake Calcasieu. Most of the sam- 38 SONIAT pling effort has been in the southern portion of the Bay. where Ray (1982) and Craig et al. (1989) also reported high prevalences. Levels of infection were particularly high in East Cove (PI = 67-l(X)'7f, WI = 1.5-3.8); Ray (1982) states that infection levels were ""among the highest I have seen in more than thirty years" study in the Gulf of Mexico." Between Lake Calcasieu and the Terrebonne Basin oyster hab- itat is scarce, and the parasite, like its host, is discontinuously distributed. Craig et al. ( 1989) report it from Joseph Harbor Bayou (PI = 73%, WI = 1.50) as well as Southwest Pass, which con- nects Vermillion Bay to the Gulf of Mexico (PI = 527f , WI = 0.33, Salinity (S) = 8 ppt). Soniat and Gauthier (1989) did not find the parasite in Southwest Pass but sampled when salinity was lower (S = 4 ppt). It is unlikely that Perkinsiis extends far if at all into Vermillion Bay. Vermillion and Atchafalaya Bays are fresh- water to oliogohaline bays that support few if any oyster popula- tions. Indeed, the salinity of the water is so depressed in this area that oyster reefs are found 5 to 6 miles offshore (Price 1954); apparently, these offshore reefs have never been sampled for "Dernio." Perkinsus has only been found near the eastern bound- ary of the Atchafalaya Basin (Craig et al. 1989). in Oyster Bayou (PI = 90<7f , WI = 0.33, S = 1 1 ppt), and Melancon et al. (1995) did not find it there when the salinity was only 1.2 ppt. Perkinsus is widespread throughout the Terrebonne/Barataria estuary. The Terrebonne Basm extends from Point au Per to Bayou Lafourche, whereas the Barataria Basin encompasses the region from Bayou Lafourche to the Mississippi River. Melancon et al. (1995) recently mapped the oyster resources of the basins and sampled for Perkinsus. The distribution of oyster and parasite in the interdistributary estuaries of Louisiana is unlike that of typical Gulf gradient estuaries. In interdistributary estuaries, most of the freshwater input is at the boundaries, with some freshwater input at the head of the estuary. In contrast, gradient estuaries receive most of their freshwater from up-estuary. As a consequence, the zone of greatest abundance of oysters in interdistributary estuaries takes the shape of an inverted "u." Lower levels of parasitism, therefore, are found not only up-estuary, but also along the pe- rimeters of the basins. In fact, one of the few populations of oysters in the Gulf of Mexico that appears to be consistently dis- ease-free is one at Tiger Pass, an outlet from the Mississippi River (Powell, personal communication). Mackin (1962) states "'areas of the lower Barataria are never free of disease" and this is prob- ably true of the adjacent lower Terrebonne Basin as well. There are, however, populations that are temporarily disease-free. For example, Melancon et al. (1995) recorded a number of uninfected samples from summer during a low-salinity period. (The highest salinity at which a negative sample was found was 11.3 ppt.) Enough studies have been conducted in the area (Mackin et al. 1950; Mackin 1951, 1953, 1956, 1962; Ray 1953, 1954a, 1954b, 1966b; Soniat and Gauthier 1989; Craig et al. 1989; Gauthier et al. 1990; Melancon et al. 1995) to conclude that Perkinsus is ubiq- uitous in and continuously distributed across the Terrebonne/ Barataria Basin. In fact, Craig et al. (1989) list Barataria Bay as one of the apparent Gulf- wide foci of infection. The Mississippi River is a natural barrier to oyster distribution; however, seed oysters from public grounds east of the Mississippi River are routinely transported to and bedded on private leases in the Terrebonne/Barataria Basin. Thus, man transports oysters and possibly infective elements from east of the Mississippi River to west of the River on an annual basis (see Ray 1996). Soniat and Gauthier ( 1989) found WI values from 0.90 to 1 .90 in seven samples taken in Louisiana east of the Mississippi River — higher than those of Mackin ( 1962), who reported a range of 0 to 1 .90. Lake Borgne is an area that supports a few low- salinity reefs (Mackin and Hopkins 1962). and populations there are intermittently infected (White et al. 1987, Soniat and Gauthier 1989, Gauthier etal. 1990). The Pearl River forms a boundary between Louisiana and Mis- sissippi. Freshwater outflow from the Pearl depresses salinities in the northern part of Lake Borgne and no oyster reefs are found there. Mackin and Hopkins (1962) indicate that there is good oyster production near Grande Isle (the one near the Louisiana and Mississippi border), but there are no records of samples taken for ""Dernio." The most extensive study of Perkinsus parasitism in Missis- sippi was that of Ogle and Flurry ( 1980). They sampled four reefs over a 25-month period (S = (V35 ppt) and found relatively low prevalences (PI = 0-609^) and intensities (WI = 0.88). Soniat and Gauthier ( 1989) report higher intensities from a single station in lower Biloxi Bay (WI = 2.00-2.27, S = 5-25 ppt). The parasite is distributed throughout the oyster-producing area of Mis- sissippi Sound and extends ar least into the lower reaches of the coastal bays (Ogle and Flurry 1980, Soniat and Gauthier 1989, Craig et al. 1989). Insufficient sampling in Mississippi makes a determination of the northern limits of the parasite in the bays difficult, but it probably coincides with the upper distribution of oyster populations. Most of the sampling effort for Perkinsus in Alabama has been in Mobile Bay, especially the southwestern portion of the Bay where the greatest concentration of reefs is found. Mackin (1962) reported very heavy WI values from several reefs, whereas Craig et al. (1989) found light infections at a single site. The most extensive sampling in the state is reported by Beckert et al. (1972). They sampled seven reefs in Mobile Bay and found the classic pattern of light infections (e.g., PI = 26.57c, WI = 0.27) in the upper part of the Bay and heavier infections in the lower Bay (e.g.. PI = 84.4%, WI = 2.03). There are no records of uninfected populations of oysters in Mobile Bay. The distribution of Perkinsus in Florida is well known. The most extensive study was that of Quick and Mackin (1971) who sampled 91 stations, 86 of which were from the northern and western coasts, and found the parasite "concomitantly with the host throughout the year and at salinities from 6 ppt to 36 ppt. " A number of studies have verified the presence of the parasite in the Pensacola Bay area (Mackin 1951, 1962; Ray 1952; Quick and Mackin 197 1; Little and Quick 1976; Craig et al. 1989). Little and Quick ( 1976) found Perkinsus in both Escambia Bay and East Bay of the Pensacola Bay estuary. Levels of disease were espe- cially high m Escambia Bay (PI = 50-100%, WI = 1.2-2.5), where Little and Quick ( 1976) ascribed the "over 90%" mortality of commercial-sized oysters in September 1971 to ""dermo dis- ease." The parasite extends into Santa Rosa Sound (Sabine Island) where it was found by Ray ( 1966a) at high prevalence (PI = 80%) and intensities (WI = 2.27). Quick and Mackin (1971 ) and Craig et al. (1989) reported Perkinsus from Choctawhatchee Bay, where the parasite is probably widespread; for example, Craig et al. (1989) found it prevalent (but not intense) at a high-salinity (S = 13 ppt, PI = 92%, WI = 0.33) site and a low-salinity (S = 0 ppt, PI = 98%, WI = 0.67) site. Data from Quick and Mackin ( 1971 ) and Craig et al. ( 1989) likewise suggest that ""Dermo" is prevalent throughout the oyster-growing areas of the St. Andrews estuary (St. Andrews Bay, West Bay, East Bay). Perkinsus marinus in the Gulf of Mexico 39 The Apalachicola Bay area typically supplies about 90% of the oysters produced in Florida (Ingle 1982). and substantial sampling for Perkinsus has been conducted there. Dawson ( 1935) sampled 401 oysters from 19 stations and demonstrated that the disease was widespread. Half of the oysters were infected; light infections occurred in 28.2%. whereas moderate to heavy infections were found in 11.0% of the oysters. Annual mean salinities at the sta- tions ranged from 7.6 to 29.3 ppt. with no difference in intensity of infection in oysters from low-salinity (<17.7 ppt) and high- salinity (>17.7 ppt) reefs. Ray (1966a) sampled three sites in the Bay— St. Vincents' Bar (PI = 0%). Cat Point Bar (PI = 100%, Wl = 2.75), and Green Point Flat (PI = 77%, WI = 1.88). Quick and Mackin (1971) found the parasite widely distributed in Apalachicola Bay and St. George Sound. Craig et al. (1989) found light infections (Wl = 0.33) at the eastern (Dry Bar. S = 13 ppt. WI = 63%) and western (Cat Point Bar. S = 7 ppt. Wl = 92%) extremes of the area. Sufficient sampling has been conducted in the region to conclude that Perkinsus is continuously distributed across Apalachicola Bay and St. George Sound. From St. George Sound to Tampa Bay, the parasite is found associated with localized oyster populations sustained by the out- flows of the Aucilla, Suwannee, Waccasassa, Crystal, and other smaller rivers (Quick and Mackin 1971 ). It has been reported from Cedar Key area by a number of investigators (Ingle and Dawson 1953. Quick and Mackin 1971. Craig et al. 1989). For example, Craig et al. (1989) found prevalences of 100% in oysters from Black Point (S = 14 ppt, Wl = 2.00). They suggest that the area east of the Mississippi River to Cedar Key. FL. has the Gulf-wide minimum for infection levels. Perkinsus is prevalent and widespread in Tampa Bay, Ray (1966a) reported it from three populations with PI values ranging from 27 to 93% and Wl values ranging from 0.40 to 1 .65. Quick and Mackin (1971) sampled 53 stations and found it throughout the Tampa Bay estuary. Craig et al. (1989) sampled four stations and found all populations 100% infected, with Wl values from 1.67 to 2.33 (S range = 28-34 ppt). The parasite has also been reported from the Charlotte Harbor area by Quick and Mackin ( 1 97 1) and Craig etal. (1989). In fact. Craig et al. ( 1989) indicate that the Tampa Bay-Charlotte Harbor area has much higher than average levels of infection and is a Gulf-wide focus of infection. Quick and Mackin ( 1971 ) report that the parasite is present near Ft. Myers (mouth of the Caloosahatchee River and behind Estero Island). Mackin (1962) indicates that a "few oysters" are pro- duced as far south as Lee County (south of Ft. Myers) and pop- ulations of C. virginica become scarce and sparse southward. Craig et al. (1989) found the parasite in Naples Bay (S = 28 ppt. PI = 100%. WI = 2.33), Rookery Bay (S = 34 ppt, PI = 100%, Wl = 0.67), and Faka Union Bay (S = 27 ppt, PI = 100%), WI = 1 .00). Quick and Mackin ( 1971 ) established sample stations for Perkinsus in the Florida Keys, but it is uncertain if the oysters sampled were C. virginica. Until further sampling extends the range, the southeastern limit of P. marinus in the Gulf of Mexico is considered to be Faka Union Bay near the Everglades of Florida (Craig et al. 1989): however. Powell (personal communication) has found Perkinsus in Bahia de Jobos, Bahia de Boqueron. and Bahia Montalva in Puerto Rico. TRANSMISSION AND PREVALENCE Environmental Factors Temperature and salinity are the main environmental factors considered in epizootiological studies of P. marinus. Although both factors and, in particular, their interaction are important (So- niat 1985), most of the variation in levels of parasitism is not explained by variations in temperature and salinity. Other envi- ronmental and biological factors, of which we know little or noth- ing, must significantly affect the levels of parasitism observed in the field. Temperature and Salinity Water temperature is one of the major environmental factors determining the prevalence and intensity of P. marinus. In some field studies (e.g.. Dawson 1955) a significant positive correlation has been established between temperature and levels of parasitism, but the relationship does not hold in all cases (e.g.. Soniat 1985). In most cases the disparity is understandable. For example, if a field survey is conducted over a short period of time, temperature varies little and may not appear to be important. Likewise, in areas where there is significant variation in salinity, salinity appears to be more important (e.g., Soniat 1985). Furthermore, in some es- tuaries temperature and salinity may be inversely correlated (e.g., Soniat 1985) and thus the two factors have opposing effects on the parasite. Recent in vitro laboratory studies (Chu and Greene 1989) have demonstrated that P. marinus is especially proliferative at temper- atures above 20°C, which corresponds well with field observa- tions. Hofmann et al. (1995) modeled the response of the parasite by assuming the effect of temperature on cell division to be a standard temperature dependency on growth. The equation takes the form TdlTj = rjlT„l c (0 oeiiMlTltl-T,,) (1) where T., = 20°C. r^lT] = the specific rate of cell division (day''), rj[T„] is the rate at 20°C, and the exponential corre- sponds to a Q,o of 2.0. AtT„ = 20°C(andS = 20ppt).rj|TJ = 0.555 day"'; that is. the population can double in size every 2 days. At 10 ppt and less, the rate of cell division is modified by salinity (Chu and Greene 1989). For salinities below 10 ppt. equa- tion ( 1 ) takes the form rjlT.SI = rjlT„.S„] (S/10) e' 0 069.M(T(O-T„) (2) where S„ = 20 ppt and r^lTJ = rjlT.,.S„l = 0.555 day ' (Ho- fmann et al. 1995). Thus, at salinities below 10 ppt the specific rate of cell division is decreased. The above equation, which quan- tifies the response to the interaction of temperature and salinity, closely tracks field data (Soniat and Powell 1994, Hofmann et al. 1995). Climatic patterns, which in turn affect local rainfall and the severity of the local winter, are related to levels of parasitism along the Gulf (Powell et al. 1992). Recent data indicate that the Gulf-wide mean infection intensity follows the El Nino cycle fairly closely (Powell, personal communication). Other Environmental Factors It has been hypothesized that pollution may affect the suscep- tibility and resistance of oysters to P. marinus (Winstead and Couch 1988. Chu and Hale 1994). Pollutants might increase dis- ease susceptibility by increasing the number or virulence of infec- tive elements. Alternatively, pollutants could reduce disease re- sistance by a physiological stress and associated energy drain on the host (see Paynter 1996) or by a suppression of the host immune system (see Anderson 1996). 40 SONIAT Winstead and Couch (1988) exposed eastern oysters to high concentrations (600 mg/1) of n-nitrosodiethylamine. resulting in significantly increased levels of P. imirimis. Chu and Hale (1994) demonstrated that water-soluble pollutants from sediments en- hanced preexisting infections and increased the oysters" suscepti- bility to infection. Field studies also suggest a link between human activity and parasite distribution. Craig et al. ( 1989) found that agricultural and urban land use affected the Gulf-wide distribution of P. marinus. Salinity and agricultural use yielded the best two-variable regres- sion model (maximum r-squared improvement), whereas the best three-variable model additionally included urban land use. (Seven variables were considered: salinity, temperature, condition index, length, residential use, industrial use, and agricultural use.) Cor- relation coefficients never exceeded 0.30. indicating that much of the variation in levels of parasitism remained unexplained. Wilson et al. (1990) developed a similar model to investigate the geo- graphical distribution of P. mahnus on a bay scale. A seven- variable model (mean latitude for each bay. agricultural land use, industnal land use. mean oyster length, total polynuclear aromatic hydrocarbons concentration |PAH|. total pesticide concentration, and salinity) explained 49% of the variation in P. inannus inten- sity. (Length is a surrogate parameter for age. which relates to pollution body burdens and levels of infection. Gulf-wide, latitude is an adequate surrogate for a temperature/salinity interaction, since the northern Gulf of Mexico is cooler and wetter than the southern Gulf.) Total PAH concentrations and industrial land use were positively correlated with prevalence, whereas latitude and agricultural land use were negatively correlated with prevalence. Biological Factors Numerous biological factors of the host potentially affect the transmission or prevalence of P mannu.s. These factors include. but are not limited to. oyster nutrition and growth, spawning and reproduction, age and resistance, and density, distribution and interactions with vectors. Nutrition and Growth Recent modeling studies (Powell et al. 1996) suggest that the timing of the spring bloom of phytoplankton is critical in the initiation of epizootics. If the spring bloom occurs early, more energy is shunted into somatic growth than into reproduction and oysters quickly accrue biomass and outgrow the parasite. The oyster and the parasite are thus in a kind of race. If the oyster can grow fast enough, the parasite concentration is effectively diluted and the oyster wins. If the growth of the oyster is slow, the parasite will win the race. An early spring bloom produces less mortality in an oyster population because more of the net production goes into somatic growth and less goes into reproduction, and one of the oyster's primary defenses is to outgrow the disease (Hofmann et al. 1992). A drop in food supply favors the parasite, since reduced food supplies do not affect the division rate of P. marinus (Powell et al. 1996) but do reduce oyster growth and fecundity (Soniat and Ray 1985). Consequently, other factors that diminish oyster growth, especially if they are operative in the summer when the parasite is rapidly proliferating, favor the initiation of epizootics in modeling simulations. Thus, high turbidity that decreases feeding efficiency, low current flow that delivers food al a suboptimal level, and dense populations of filter feeders that outcompete the oyster for food, all reduce oyster growth rates and. in madi'ling scenarios if not nature herself, favor the parasite over the host (Powell et al. 1996). Spawning and Reproduction Early observations suggested that heavy mortalities of oysters often followed spawning. It was hypothesized that the stress of spawning weakened host oysters such that they were more easily infected and had less resistance to the progression of disease. Mackin (1953) tested this hypothesis, and although he found no relationship between levels of infection in pre- and postspawning oysters, he observed a degeneration of the gonads if the disease struck in an early stage of gonadal development (see also Mackin 1962). Subsequent studies by Wilson et al. (1988) and Choi et al. (1993, 1994) suggest that P. marinus can decrease oyster fecun- dity. Choi et al. ( 1994), for example, found a relationship between the rate of gamete production and infection intensity in certain months. Heavy mortalities of oysters following spawning have an alternate explanation. Spawning reduces oyster biomass more than "Dermo" biomass so that the number of cells per gram of oyster increases, possibly to lethal levels. The effect of spawning success on disease progression in an oyster population may be of even more significance than the effect of level of disease on oyster fecundity. For example, Powell et al. (1996) suggest that recruitment failure is likely a principal mech- anism increasing population infection intensity and initiating an epizootic. Likewise, one way in which an epizootic can be termi- nated IS by a massive, successful spawning event. New and un- infected biomass thus replaces old, infected oysters that have died from the disease. Infection intensity in juvenile oysters is reduced, and if enough food is present, juvenile oysters grow fast enough to dilute P. marinus in the population and maintain population fe- cundity (Powell et al. 1996). Age and Resistance Mackin ( 145 1 ) reported that Louisiana oysters less than 1 year ot age are not infected as extensively as are market-sized oysters. Ray ( 1953). also working in Louisiana, found that spat held in an area of high endemism appeared to be highly refractive to infec- tions. Andrews and Hewatt (1957). however, fed a tissue mince of heavily infected gapers to young oysters which became infected and died as quickly as older control oysters. Subsequent studies in Texas (Hofstetter 1977. Ray 1987) suggested that young oysters under natural conditions could become heavily infected at an early age. Apparently then, there is no inherently greater immunity in young as compared to old oysters. The observation that young oysters often have lower levels of infection can be explained by the fact that they must be exposed to a sufficient number of infective elements to initiate an infection. Furthermore, young oysters are growing rapidly when the parasite is in its initial '"lag phase"" of growth before Perl^insus population growth is greatly accelerated. Thus, if enough infective elements are present in the environment, young oysters can become heavily infected, especially if the pro- liferation of the parasite population is rapid relative to the growth of the oyster host. When oysters are young, their cell division rate approximates the cell division rate of ""Dermo." Thus, under op- timal conditions of growth for the host, the parasite cannot grow fast enough to reach a high intensity until the oyster reaches adult- hood, when its growth rate declines. Density, Distribution, and Vectors The close proximity of infected oysters to uninfected oysters appears to be an important factor in the spread of P. marinus (Ray Perkinsus marinus in the Gulf of Mexico 41 19871. The potential for an epizootic is thus related to the distance of the uninfected population from a source of infection, the pre- vailing water currents, and possibly the efficacy of transmission by vectors — all of which determine the number of infective elements available to the susceptible population. Transmission, however, includes two distinct processes which operate on different spatial and temporal scales — reef-to-reef transmission which occurs over relatively greater distances and times, versus oyster-to-oyster transmission on reefs which acts over shorter distances and re- quires less time. Andrews (1988) reviewed the results of a number of tray ex- periments designed to test the effectiveness of transmission over varying distances. His conclusion was that transmission of P. marinus tends to be localized and that isolation of beds is a useful management strategy for controlling the disease during normal years (see also Andrews and Ray 1988). The effects of isolation in retarding the spread of the disease are also supported by studies of Ray (1987) in Galveston Bay. TX. Of particular importance is the occurrence of "killing floods" which destroy oyster populations, diminish infection levels, and ultimately decrease the number of infective elements in the estuary. Such events provide "natural experiments" in which the temporal and spatial response of the system can be observed. Freshet mortalities remove diseased oys- ters from the estuary, provide new, clean shell for spat settlement, and "reset" the system at a lower level of infection. The success of the system response (again) depends initially upon the success of oyster recruitment and subsequently upon the ability of the oysters to outgrow the disease (Powell et al. 1996). High recruit- ment and fast growth help explain why Gulf oyster populations thrive at temperatures and salinities that would cause the demise of their northern counterparts. Although the evidence clearly suggests that the direct transmis- sion of waterbome infective elements from oyster to oyster is the primary mechanism for spreading the disease (Ray 1954a, 1954b; Mackin 1962; Perkins and Menzel 1966; Andrews 1988), scaven- gers may also be important as disease vectors (Ray 1954b, Hoese 1962, White et al. 1987). Numerous scavengers live on Gulf oyster reefs and consume dead and dying oysters. They include, among others, blue crabs, mud or xanthid crabs, nereid poly- chaetes. and fishes such as gobies and blennies (Andrews 1988). Hoese (1962), for example, found live Perkinsiis in the intestinal tracts of oyster drills (Urosalpiitx cinerea) and fishes (Gobiosoma bosci, Chasmodes bosquianus, Opsanus tau) and on the bodies of xanthid crabs (Neopanope lexana. Rithropanopeus harrisii). Gap- ers (dead oysters with the meat intact) are rarely observed in the field, and it is likely that many of the Perkinsiis cells from dead oysters pass through the digestive systems of scavengers before the oysters decay (Hoese 1962, Ray 1987). White el al. (1987) demonstrated that the ectoparasitic snail. Boonea ( = Oduslomia) impressa. directly transfers Perkinsiis from uninfected to infected oysters. These ectoparasites puncture the mantle edge of oysters to suck body fluids and transfer viable Perkinsiis cells as they move from oyster to oyster. SUMMARY AND CONCLUSIONS Although P. marinus is one of the most studied marine patho- gens, much remains to be learned. Its distribution in the Gulf of Mexico is fairly well known; exceptions are Mexico where it has not been extensively sampled and some bays where its upper limit has not been established. Its relationship with temperature and salinity is well documented, yet these factors do not explain most of the variation in levels of parasitism observed in the field. A more complete accounting for observed patterns will require a more complex explanation. Of likely importance will be the in- teraction of temperature and salinity, the timing of the spring bloom in relation to the increase in temperature in the spring, and the interplay of physical factors (especially temperature and salin- ity), nutritional factors, and host health and condition. To achieve such an understanding requires a model of the disease process, field and laboratory verification of model responses, and ongoing iterative interactions among model, field, and laboratory studies. ACKNOWLEDGMENTS 1 am indebted to mentors and colleagues who have greatly influenced my thinking concerning Perkinsiis — in particular, Drs. Sammy M. Ray, John G. Mackin, Eric N. Powell, Frank O. Perkins, and Fu-Lin Chu, Dr. Henry Hildebrand graciously pro- vided me with background on the history of studies of P. marinus in Mexico. The manuscript was greatly improved by the comments of two anonymous reviewers. Special thanks to Diedre Gibson, Randy Robichaux. and Darrell Solet for their assistance in pro- ducing the map of the Gulf of Mexico. LITERATU Anderson, R. S. 1996. Interactions of Perkinsiis iminniis with humoral factors and hemocyles of Crassostrea virginica. J. Shellfish Res 15:127-134 Andrews, J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsiis marinus and its effects on the oyster industry. Amer. Fish. Soc. Spec. Piibl. 18:47-63. Andrews, J. D. & W. G. Hewatt. 1957. Oyster mortality studies in Vir- ginia II. The fungus disease caused by Dermocyslidiwn mariniim in oysters of Chesapeake Bay. Ecol. Monogr. 27:1-26. Andrews, J. D. & S. M, Ray. 1988. Management strategies to control the disease caused by Perkinsiis niarinii.s. Amer. Fish. Soc. Spec. Pitbl. 18:257-264. Beckert, H., D. G. Bland & E. B. May. 1972. The incidence of Laby- rinthomy.xa marina in Alabama. Ala. Mar. Resoiir. Bull. 8:18-24. Burreson, E. M., R. S. Alvarez, V. V. Martinez & L. A. Macedo. 1994. Perkinsiis marinus ( Apicomplexa) as a potential source of oyster Cras- sostrea virginica mortality in coastal lagoons of Tabasco, Mexico. Dis. Aqiiat. Org. 20:77-82. Choi. K-S., D. H. Lewis, E. N. Powell & S. M. Ray, 1993. Quantitative measurement of reproductive output in the American oyster. Crassos- RE CITED irea virginica. using an enzyme-linked immunosorbent assay (ELISA). Aqiiaciilt. Fish. Mgmt. 24:299-322. Choi, K-S., E. N. Powell, D. H. Lewis & S. M. Ray. 1994. Measure- ments of instantaneous reproductive effort in the American oyster, Crassostrea virginica. using a Protein A immunoprecipitation assay. Biol. Bull. 186:41-61. Chu, F-L. E. & K. H. Greene. 1989. Effect of temperature and salinity on the in vitro cuhure of the oyster pathogen Perkmsus marinus (Apicom- plexa: Perkinsea). J Invertebr. Pathol. 53:260-268. Chu, F-L. E. & R. C. Hale. 1994. Relationship between pollution and susceptibility to Infectious disease in the eastern oyster. Crassostrea virginica. Mar. Environ. Res. 38:2243-2256. Craig. A.. E. N. Powell, R. R. Fay & J, M. Brooks. 1989. Distnbution of Perkinsus marinus in Gulf Coast oyster populations. Estuaries 12:82- 91. Dawson, C. E. 1955. Observations on the intensity of Dermocyslidium marinum infection In oysters from Apalachicola Bay, Florida. Tex. J . Sci. 7:47-56. Gauthier, J. D. & W. S. Fisher. 1990 Hcmolymph assay for diagnosis of 42 SONIAT Perkinsus marimis in oysters. Crassostrea virginica (Gmelin 1791 ). J . Shellfish Res. 9:367-371. Gauthier. J. D., T. M. Soniat & J. S. Rogers. 1990. A parasitological survey of oysters along salinity gradients in coastal Louisiana. J . World Aqitacult. Soc. 21:105-115. Groue, K. J. & L. J. Lester. 1982. A morphological and genetic analysis of geographic variation among oysters m the Gulf of Mexico. Veliger 24:331-335. Hoese, H. D. 1962. Studies on oyster scavengers and their relation to the fungus Dermocystidiiim marinum Proc. Natl. Shellfish Assoc. 53: 161-174. Hoese, H. D. 1963. Absence oi Dermocyslidiwn marinum at Port Aran- sas, Texas, with note on an apparent inhibitor. Te.x. J. Sci. 15:98-103. Hofmann, E.E..E.N. Powell. J. M. Klink&G. Saunders. 1995. Modeling diseased oyster populations. I. Modeling Perkinsus marnuis infections in oysters. J. Shellfish Res. 14:121-151. Hofmann, E. E.,E. N.Powell. J. M.Klink&E. A. Wilson. 1992. Mod- eling oyster populations. IlL Critical feeding periods, growth and re- production. J. Shellfish Res. 11:399-416, Hofstetter, R. P. 1977, Trends in population levels of the American oys- ter, Crassostrea virginica (Gmelin) on public reefs in Galveston Bay, Tex. Parks Wildl. Depl.. Tech Ser. No. 24., 90 p. Ingle, R. M. 1982. The Florida oyster fishery: an update. In: K. K. Chew (ed.). Proceedings of the North American Oyster Workshop. L.S.U., Div. Cont. Ed., Baton Rouge, Louisiana, pp. 116-119. Ingle, R. M. & C. E. Dawson. 1953. A survey of the Cedar Key area. Fla. St. Bd. C onsen-. Tech. Ser. No, 9, Levine, N, D. 1978. Perkinsus gen. n. and other new taxa in the protozoan phylum Apicomplexa. J. Parasitol. 64:549. Little, E. J. & J. A. Quick Jr. 1976, Ecology, resource rehabilitation and fungal parasitology of commercial oysters, Crassostrea virginica (Gmelin) in Pensacola estuary. Florida, Fla. Dept. Nat. Resaur. Fla. Mar. Res. Piibl. 21:1-89. Mackin, J. G. 1951. Histopathology of infection oi Crassostrea virginica (Gmelin) by Dermocystidium marinum Mackin, Owen and Collier. Bull. Mar. Sci. 1:72-87. Mackin, J. G. 1953. Incidence of infection of oysters by Dermocystidium in the Barataria Bay area of Louisiana. Natl. Shellfish. Assoc. Conv. Add. 1951:22-35. Mackin, J. G. 1956. Dermocystidium marinum and salinity, Proc. Nail. Shellfish. Assoc. 46:116-128. Mackin, J. G. 1962. Oyster diseases caused by Dermocystidium marinum and other microorganisms in Louisiana, Puhl. Inst. Mar. Sci. Univ. Tex. 7:132-229. Mackin, J. G. & J. L. Boswell. 1956. The life cycle and relationships of Dermocystidium marinum. Proc. Natl. Shellfish. Assoc. 46:112-115. Mackin. J. G. & S. H. Hopkins. 1958. Results of Projects Nine and Twenty-Three: a summary report. Tex. A&M Res. Found. Coll. Sta- tion. Tex. 200 p. Mackin, J. G. & S. H. Hopkins. 1962, Studies on oyster mortality in relation to natural environments and to oil fields in Louisiana. Puhl. Inst. Mar. Sci. Univ. Tex. 7:1-131. Mackin, J. G., H. M. Owen & A. Collier. 1950. Preliminary note on the occurrence of a new protistian parasite, Dermocystidium marinum n sp. in Crassostrea virginica (Gmelin). Science 111:328-329 Mackin. J. G. & S. M. Ray. 1966. The taxonomic relationship of Der- mocystidium marinum Mackin, Owen and Collier. J. Iiiveriehr. Pathol. 8:544-545. Mackin, J. G. & A. K. Sparks. 1962. A study of the effect on oysters of crude oil loss from a wild well. Cont. Mar. Sci. 7:230-261. Melancon, E. Jr., T. M. Soniat, V. Cheramie. R. Dugas & J. Barras. 1995. Oyster resource zones within Louisiana's Barataria and Terre- bonne estuaries. Final report to the Barataria/Terrebonne National Es- tuary Program. 155 p. Ogle, J. & K. Flurry. 1980. Occurrence and seasonality of Perkinsus marinus (Protozoa: Apicomplexa) in Mississippi oysters. Gulf Res. Repl. 6:423-t25. Paynter, K. T, 1996, The effects of Perkinsus marinus infection on phys- iological processes of the eastern oyster Crassostrea virginica. J. Shellfish Res. 15:119-125. Perkins, F. O. 1969. Ultrastructure of vegetative stages in Lahyrithomv.xa marina ( = Dermocystidium marinum), a commercially significant oys- ter pathogen, J. Invertehr. Pathol. 13:199-222. Perkins, F. O. 1976. Zoospores of the oyster pathogen, Dermocystidium marinum. I, Fine structure of the conoid and other sporozoan-like organelles. J. Parasitol. 62:959-974. Perkins, F. O. 1996. The structure of Perkinsus marinus (Mackin. Owen & Collier, 1950) Levine 1978 with comments on taxonomy and phy- logeny of Perkinsus spp. J. Shellfish Res. 15:67-87. Perkins. F. O. & R, W, Menzel. 1966, Morphological and cultural stages in the life cycle of Dermocystidium marinum. Proc. Natl. Shellfish. Assoc. 56:23-30, Powell. E. N., J. D. Gauthier, E. A. Wilson, A. Nelson, R. R. Fay & J. M. Brooks. 1992. Oyster disease and climate change. Are yearly changes in Perkinsus marinus parasitism in oysters (Crassostrea vir- ginica) controlled by climatic cycles in the Gulf of Mexico? P.S.Z.N.I.: Mar. Ecol. 13:243-270. Powell. E. N., J. M. Klinck, E. E. Hofmann & S. M. Ray, 1994. Mod- eling oyster populations. IV: rates of mortality, population crashes, and management. Fish. Bull. 92:347-373. Powell. E, N,. J. M. Klink cS: E. E. Hofmann. 1996. Modeling diseased oyster populations II. Triggering mechanisms for Perkinsus marinus epizootics. J. Shellfish Res. 15:141-165. Pnce. W. A. 1954. Oyster reefs of the Gulf of Mexico. In: P. S. Galtsoff (ed.). Gulf of Mexico: Its Origin. Waters and Marine Life. U.S. Fish and Wildlife Ser,. Fi.sh. Bull, 89:491, Quick. J, A. & J. G. Mackin, 1971 , Oyster parasitism by Labrinlhomy.xa marina in Florida, Fla Depl. Nal. Resour. Mar. Res. Lab. Prof. Pap. Ser. 13:1-55. Ray, S. M. 1952. A culture technique for the diagnosis of infections with Dermocystidium marinum Mackin. Owen, and Collier in oysters. Sci- ence 116:360-361, Ray, S. M, 1953. Studies on the occurrence oi Dermocystidium marinum in young oysters, Proc. Natl. Shellfish. Assoc. 44:80-92. Ray, S. M. 1954a. Experimental studies on the transmission and patho- genicity of Dermocystidium marinum. a fungous parasite of oysters. J. Parasitol. 40:235. Ray, S. M. 1954b. Biological studies of Dermocystidium marinum. a fun- gus parasite of oysters. Rice Inst. Pamph. Spec. Issue 1 14 p. Ray. S. M. 1966a, Notes on the occurtence of Dermocystidium marinum on the Gulf of Mexico dunng 1961 and 1962. Proc. Natl. Shcllfi.sh. Assoc. 54:45-54, Ray, S. M. 1966b. A review of the culture method for detecting Dermo- cystidium marinum. with suggested modifications and precautions. Proc. Natl. Shellfish. Assoc. 54:55-69. Ray. S M. 1982. Summary report on monitoring program in Lake Cal- casieu. Louisiana. Ray Biol. Co.. Galveston. Texas 35 p. Ray. S. M, 1987, Salinity requirements for the American oyster. Cras- sostrea virginica. In: A, J, Mueller & G, A. Matthews (eds.). Fresh- water Inflow Needs of the Matagorda Bay System With Focus on Penaeid Shrimp. U.S. Dept. of Commerce, NOAA Tech. Mem, NMFS-SEFC-189, pp. E1-E28. Ray. S. M. 1996. Historical perspective on Perkinsus marinus disease of oysters in the Gulf of Mexico. J. Shellfish Res. 15:9-1 1. Ray, S. M., J. G. Mackin & J. L. Boswell. 1953. Quantitative measure- ment of the effect on oysters of disease caused by Dermocystidium marinum. Bull. Mar. Sci. 3:6-33. Soniat, T. M. 1985. Changes in levels of infection of oysters by Perkinsus marinus. with special reference to the interaction of temperature and salinity upon parasitism. N.E. Gulf Sci. 7:171-174. Soniat, T. M. & J, D. Gauthier. 1989. The prevalence and intensity of Perkinsus marinus from the mid northern Gulf of Mexico, with com- ments on the relationship of the oyster parasite to temperature and salinity. Tul. Stud. Zool. Bot. 27:21-27. Soniat, T, M. & E. N. Powell. 1994, The effects of temperature, salinity PhRKI,\SL'S MAR/M'S IN THE GULI- OF MllXICO 43 and food supply on oyster production in Louisiana, model prcdiLtK)ns versus field data. 7. Shellfish Res. 13:290. Soniat, T. M. & S. M. Ray. 1985. Relationships between possible avail- able food and the composition, condition and reproductive state of oysters from Galveston Bay. Te.xas. Coiit. Mar. Sci. 28:109-121. Soniat. T. M. L. E. Smith & M. S. Brody. 1984. Mortality and condition of oysters in Galveston Bay. Texas. Com. Mar. Sci. 51:77-94. White. M. E.. E. N. Powell. S. M, Ray & E. A. Wilson. 1987. Host-to- host transmission of Perkinsiis muninis in oyster {Crcissostrea \irf>in- ica] populations by the ectoparasitic snail Botineu impressa (Pyra- midellidae). J. Shellfish Res. 6:1-5. Wilson. E. A . E. N. Powell. M. A. Craig. T. L Wade & J. M. Brooks. 1990. The distribution oi ferkiiisiis inurnms in Gulf Coast oysters: its relationship with temperature, reproduction, and pollution body bur- den. Inl. Revue Ges Hydrohinl. 75:5.V1-550. Wilson, E. A.. E. N. Powell & S. M. Ray. 198X. The elfect of the ec- toparasitic pyramidellid snail, Boonea impressa. on the growth and health of oysters. Crassoslrea virginica. under field conditions. Fish. Bull 86:553-566. Winstead, J. T. & J. A. Couch. 1988. Enhancement of protozoan patho- gen Perkinsus mariniis infections in American oysters Crassoslrea vir- ginica exposed to the chemical carcinogen «-nitrosodiethylamine (DENA). Dis Aqiiai Ori;. 5:205-213. Jcntnuil of Shellfish Rcsecinh. Vol. 15. No. 1,45-56. 1996. RANGE EXTENSION BY THE OYSTER PARASITE PERKINSUS MARINUS INTO THE NORTHEASTERN UNITED STATES: RESPONSE TO CLIMATE CHANGE? SUSAN E. FORD Rutgers Universit}' Institute of Marine and Coastal Sciences and New Jersey Agricultural Experiment Station Haskin Shellfish Research Laboratory RD#1 Box B-8 Port Norris. New Jersey 08349 ABSTRACT From its discoverv in 1949 unlil 1990. the oyster parasite Pcrkiiisiis inaniuis. cause of Dermo disease in the eastern oyster Cnissostreci virginica. was found primarily from Chesapeake Bay south along the Atlantic Coast of the United States and into theGulf of Mexico. In 1990 and 1991. the parasite suddenly appeared in locations from Delaware Bay. NJ, to Cape Cod. MA. a range extending more than 500 km north of Chesapeake Bay. An earlier incursion of the parasite into Delaware Bay in the 1950s, associated with importation of large numbers of infected oysters from Chesapeake Bay. did not cause detectable mortalities or result in the establishment of a significant parasite population. The parasite was no longer detected after imports of infected oysters ceased. In contrast, the epizootic that began in Delaware Bay in 1990 resulted in high disease prevalence and intensity, and caused heavy mortalities, but was not linked (o similar imports. Several hypotheses for the sudden appearance of the parasite in the northeastern United States are considered: 1 ) the parasite was transmitted via infected oysters introduced from enzootic southern areas into northern waters; 2) a change in the genetic structure of either host or parasite increased the parasite's ability to invade and proliferate in the northeast; 3) the environment in the northeastern United States became more favorable for parasite activity; or 4) some combination of these three. The .simplest explanation consistent with available data is that the pathogen was repeatedly introduced, by many means over many years, into various northeast locations where it remained undetected and was stimulated to proliferate into an epizootic by a recent extreme warming trend. Above average winter, rather than summer, temperatures were associated with the 1990s epizootic. Also, cold winters, not cool summers, were correlated with the disappearance of P. marinus from Delaware Bay in the 1950s. Stopping or materially slowing the epizootic will probably require a series of consecutive cold (i.e., average or below average temperatures) winters and cool springs that will delay and restrict the proliferation of parasites during the following summer. Eliminating the parasite from its new range may be difficult even with cooler temperatures, however, as the development of low temperature-adapted parasites could occur now that large populations are established in a region where selection pressure exists for this trait. KEY WORDS: Dermo disease. Crassostrea virginica. temperature, genetic change, parasite introduction, environmental change HISTORICAL DISTRIBUTION OF PERKINSUS MARINUS In 1940. oyster growers in Louisiana began reporting unusual mortalities of oysters and blamed oil pollution caused by compa- nies drilling in the Gulf of Mexico (Mackin and Sparks 1962). Scientists hired by the oil companies in 1947 soon identified a protozoan. Perkinsus i. = Dennocysndnim. = Lcilnrinrhomyxa) marinus. as the cause (Mackin et al. 1950. Mackin and Hopkins 1962). At about the same time, researchers in Virginia found the same parasite in oysters that had sui^'ived a 1949 mortality in the Rappahannock River (Andrews 1955). Over the next few years, P nniruuis was found in oysters throughout the southeastern United States and Gulf of Mexico (Ray 1954). The disease caused by P. marinus is commonly called "Dermo" disease, a reference to the genus (Dermocystidium) in which the parasite was originally placed. Although P. marinus was found over a wide area during a relatively brief period by scientists looking for it, the pathogen had likely been present for many years. Mackin and Sparks (1962) searched published records of the oyster industry in Louisiana dating back to the early 1900s and found that reported mortalities had the characteristics later found to describe P. marinus activity (see Andrews 1988). The mortalities occurred during warm, dry periods; they selectively affected older oysters: and they did not affect other members of the oyster community. P. marinus was also identified in archived tissue sections of Louisiana ovsters fixed around 1930 (Owen pers. comm. in Ray 1954). Andrews and Hcwatt (1957) considered it probable that the parasite had existed in Chesapeake Bay before they first recorded it in 1949. This belief was supported by the fact that the Virginia industry harvested annually as many oysters between 1950 and 1959, years of high disease levels, as it had during the two decades before P. maruuis was discovered (Haven et al. 1978). CHANGES IN THE DISTRIBITION OF P. MARINUS IN THE 1980s AND 1990s With the exception of a localized and nondestructive incursion into Delaware Bay, described below, and findings in the upper Chesapeake Bay in the mid-1970s (Otto and Krantz 1977), detect- able P marinus activity remained limited to waters from the lower Chesapeake Bay south. Reports of the parasite in Connecticut and Massachusetts (Sindermann 1970, Quick 1977) are given without documentation. In the 1980s, this observed distribution began to change (Fig. 1 ). From 1985 to 1988, P. marinus was detected, and caused mortalities, increasingly farther upestuary in the Maryland portion of Chesapeake Bay (Burreson and Ragone Calvo 1996). It appeared, about the same time, in the small estuaries along the Atlantic Coast from Virginia to southern New Jersey (Table I; Burreson and Ragone Calvo 1996). In 1990. a major epizootic began on the New Jersey side of Delaware Bay. From 1991 to 45 46 Ford 44.5- 42 5- 40.5- 38.5- 36.5- Marlha"s Vineyard, Massachusetts 1994 Southern Massachusetts 1991 78 Washington Long Island Sound 1992 Northern New Jersey Coast 1991 Southern New Jersey Coast 1 982 Delaware Bay 1990 Jppei Chesapeake Bay 1985-1988 — Seaside ol Virginia and Maryland 1985-1988 -Lower Chesapeake Bay 19-49 ATLANTIC OCEAN I 70 I 68 66 Figure I . Map of the Atlantic Coast of the northeastern I'nited States showing the dates when P. marinus was first detected. The three locations depicted with an encircled X are stations from which long-term air temperature deviation records were analyzed (see Fig. 6). 1992, the parasite was found progressively northward along the cursions of the parasite into Delaware Bay. documents the ob- Atlantic Coast to Cape Cod. MA. In 1995, its presence was con- served spread of P. marinus northward from Chesapeake Bay firmed in Maine. beginning in 1990, and discusses various hypotheses for its sudden This paper compares and contrasts the 1950s" and 1990s" in- range extension. TABLE 1. History of P. marinus in New Jersey oysters. Prevalences are the maxima recorded, usually in autumn. Note that the extent of sampling and the type of diagnosis varied from year to year. Year '7t Prevalence Comments Delaware Bay planters begin importing oyster seed from lower Chesapeake Bay A tew infected oysters detected in lower Delaware Bay (Andrews & Hewatt 1957) Regular RFTM monitoring in Delaware Bay initiated by Rutgers Oyster Research Lab Regular RFTM monitoring in Delaware Bay First outbreak of MSX disease on New Jersey leased grounds Mortalities from MSX disease spread throughout bay Embargo on all oyster imports into and exports from Delaware Bay Regular RFTM monitoring of Delaware Bay Regular RFTM monitoring of Delaware Bay Regular RFTM monilonng of Delaware Bay Regular RFTM monitoring of Delaware Bay Regular RFTM monitoring discontinued except at Cape Shore, where P marinus still present Infections found in tissue sections from Maurice River Cove grounds and followed up with RFTM culture; some gapers heavy; nothing found at other locations tested Heavy infections found in early July in lower seed bed oysters after < 1 month at Cape Shore No regular RFTM monitoring except at Cape Shore, where P. marinus still present Infected oysters no longer found at Cape Shore Infections found in tissue sections of oysters from coastal New Jersey bays and rivers (Great Bay, Tuckahoe River, Great Egg Harbor Bay) One infected oyster found in tissue sections of 20 oysters from lower New Jersey seed beds Very light infections found on Delaware Bay natural oyster beds off Kelly Island, DE, and Egg Island Point. NJ (Hofmann et al. 1945) Atypical mortalities in Maunce River and at Cape Shore signal onset of epizootic and resumption of RFTM sampling Intensification and spread on New Jersey side of Delaware Bay; few or no infections on the Delaware side 1952 1953-54 4 1955 40-60 1956 10-15 1957 25-35 1958 20-25 1959 8 I960 8 1961 0 1962 4 1963 0 1964-73 1975 52 1976 1975-80 1981 1982-87 10-14 1985 5 1988 5 1990 90 1991-95 100 PERKf\Si'S MARI\L'S RaNGE EXTENSION 47 INCURSIONS OF P. MARISLS INTO DELAWARE BAY: 195(ls AND 1990s The 1950s Epizootic The oyster industries of both Delaware and New Jersey devel- oped and now operate in similar fashions (Ford 1996). Seed oys- ters are transplanted each year from the public beds in the upper bay onto private leased grounds in the lower bay (Fig. 2), where they grow and fatten until they reach market quality. Historically, when native seed was insufficient to supply the needs of planters, oysters were brought in from other areas. Seed oysters from Ches- apeake Bay were imported into Delaware Bay for decades begin- ning in the early 1880s (Goode 1887 quoted in Andrews and Hewatt 1957. Ford 1996). but the practice appears to have ex- panded around 1950 when the native seed supply was severely depleted (H. Bickings. Sr.. Bivalve Packing Co.. pers. comm. 1989). Originally, most of the seed came from the James River. where P mariniis was not present, but in 1952-53 it became illegal to sell James River oysters for direct shipment out of Vir- ginia (.Andrews and Hewatt 1957). Thereafter, many Delaware Bay planters bought "seed" oysters from private leases in the Hampton Roads area and other higher salinity regions of Chesa- peake Bay where P. maruuis was present and causing heavy losses (Andrews 1988). Infected oysters were brought by the shipload for planting on the leased grounds of lower Delaware Bay. From the studies of Mackin (1962) and Ray ( 1954) in the Gulf of Mexico, and Andrews and Hewatt ( 1957) in Chesapeake Bay. it was already clear that P nuininis could be transmitted directly from oyster to oyster. Limited surveys had detected a few infected oysters in Delaware Bay in 1953 and 1954. and some planters had reported losses in 1954 and 1955 (G. Christensen 1956. unpub- lished observation). To determine whether the disease had spread from the imported oysters, a survey was begun by the Rutgers University Oyster (now Haskin Shellfish) Research Laboratory, under the direction of Dr. Harold Haskin. Extensive sampling was 39 7. 39 5. 39,3- 39 1- 38.9. 38.7. 75 7 ATLANTIC OCEAN 75 3 75 1 74 9 performed throughout the year over a 4-year period from 1955 through 1958. first by Ms. Greta Christensen and later by Mr. Donald Kunkle. Collections were made from all regions of the New Jersey side of Delaware Bay. including both leased grounds and seed beds. Less extensive sampling, concentrated on the leased grounds in late summer and autumn, was continued until 1963. From 1955 to 1958. particular attention was paid to sam- pling the imported Virginia seed and adjacent native oysters. Oys- ters were processed using the standard Ray's fluid thioglycollate medium (RFTM) assay to detect P. marinus and the Ray/Mackin rating system to estimate sample weighted prevalence on a scale from 0 to 5 (Ray 1954). Infected oysters were found almost exclusively from late sum- mer into early winter; detectable infections were few or nonexis- tent from late winter through early summary (Fig. 3). Peak prev- alence, recorded during the late summer and autumn, reached 50-609( on a few grounds, but generally did not exceed 40%. Most infections were in oysters brought from Virginia and in the native oysters planted nearby (Fig. 4A); however, disease levels were not consistently different between the two groups. Infections were negligible on the seed beds and in the far eastern edge of the leased grounds where oysters were rarely planted. The highest weighted prevalences were between 1.0 and 1.3. These scores were low compared to those in Virginia and the Gulf of Mexico, but a rating of 1 .0 or more is sufficient to cause some deaths in an affected population (Andrews and Hewatt 1957. Mackin 1962). Mortalities were not assessed during the Rutgers study, but based on weighted prevalences, it was estimated that losses due to P . marinus could not have exceeded 20% annually and were only about 5% in most areas (G. Christensen 1956. unpublished obser- vation). The localized distribution of P. imiriiius and the proximity of infected native oysters to imported stocks were highly suggestive of transmission from the Virginia oysters. Lack of preintroduction monitoring, however, prevented unequivocal determination of the parasite's origin. A few years after the introductions had begun, another event occurred that provided additional evidence that the Virginia oysters had introduced P. marinus into Delaware Bay. In the spring of 1957. heavy mortalities of oysters began on the New Jersey leased grounds. By 1959. they had spread throughout the 100 O Q- < (/5 O 2 Q >02aDQ:a:>z 2 • I I ^e3Q-*->ozm:cQ:>z=JOQ-*->ozmtrcc ^DLUOOLiJCTlO^(J) Figure 6. Cumulative monthly air temperature deviations at three stations in the northeastern United States adjacent to areas where P. marinus has become epizootic since 1990 (see Fig. 1). Plots were ob- tained by summing the difference (positive or negative) between monthly air temperatures and the long-term mean (1951-80) for that month. A negative slope indicates a period during which a string of months had below average temperatures; a positive slope indicates a period during which a string of months had above average tempera- tures: and a line without a clear slope represents a period of near average temperatures. The steep positive slope outlined in each plot represents a 2-year extreme warming period common to all three sta- tions. (January-March; Fig. 7A). Just after the large-scale importation of infected oysters into the bay began in the 1950s, a prolonged series of winters with below average temperatures also began. A shorter series of above average temperature winters in the early and mid- 1970s culminated in the localized P. marinus outbreak in 1975 described earlier. The series of cold winters beginning in 1977 was associated with the disappearance of detectable P . marinus infec- tions in trays of experimental oysters at the Rutgers Cape Shore Laboratory. The occasional findings in Delaware Bay in the mid- 1980s were at a time when winter temperatures were either some- 54 Ford 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 1.5 CO « 05H LU O CO 0 LJJ LU g-0,5 LU Q -1 -1,5 g SUMMER (JULY-SEPTEMBER) I ^ jl iiii 111 i|irr JLL ■I i|r|rriiri I'l I.II I Ill 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 YEAR Figure 7. Seasonal air temperature deviations at Milhille. NJ, from 1949 tlirough 1994. Deviations are the difference between mean tem- peratures for each 3-month period for each year of record and the long-term mean for that period. what above, or only slightly below, average and when there was also a string of above average summer temperatures (Fig. 7B). Finally, the current epizootic is correlated with four above average temperature winters beginning in 1989. No such pattern emerged from plots of any other season, including July-September when P. marinus is most active in Delaware Bay (Fig. 7B). Warm winters would act by lessening overwinter parasite mor- tality (Bushek et al. 1994, Ragone Calvo and Burreson 1994) and accelerating the development of advanced, lethal infections in the late spring. Release of infective particles from heavily infected and dying oysters early in the season would then lead to a new round of infections resulting in multiple infection cycles in a single sum- mer (Andrews 1988). SUMMARY AND CONCLUSIONS The 1990-91 observed range extension by P . marinus mto the northeastern United States can be linked empirically to historical, rather than new, introductions of potentially infected oysters and a recent warming trend that has made the environment suitable for parasite development. The possibility that P . marinus has become more cold adapted has not been tested and cannot be excluded as an additional factor. Some selection for cold tolerance may have been occurring gradually, particularly during brief periods of above average temperature when heightened parasite multiplica- tion rates would have increased the parasite population on which selection could act. The 1950s Delaware Bay episode clearly demonstrates that the mtroduction of a disease agent — even in large quantity — is not, by iisi'If. sufficient to cause a damaging outbreak of that disease or the establishment of a significant population of the agent. To become well established and then to trigger an epizootic in the new loca- tion, the introduced pathogen requires that certain conditions be favorable and that they last long enough for the population to attain a threshold level. Yet, even if these conditions are not met, small numbers of parasites may be able to survive in a generally unfa- vorable environment. Failure to detect a pathogen is not sufficient evidence that the organism is absent because population abun- dance may be below a "threshold of perception" (Fig. 8). The threshold is reached by the appearance of dead and dying hosts if "perception" is defined as morbidity and mortality. For example, had there not been a specific monitoring program for P. marinus in Delaware Bay. its presence would not have become evident until 1990. when mortalities began. If diagnostic methods that detect nonlethal infections are in routine use, the perception threshold is much lower and detected infections may not signal an imminent epizootic. Even these, however, may be too insensitive to detect very low pathogen densities. It is relevant to the follow- ing argument that the standard Ray/Mackin assay does not reliably detect very light P. marinus infections. A positive diagnosis typ- ically requires total-body parasite densities greater than 10' g" ' wet weight (Choi et al. 1989. Bushek et al. 1994). The simplest explanation consistent with the available data for the recent observed range extension by P. marinus is that the pathogen was repeatedly introduced, by many means over many years, into various northeast locations where it persisted at unde- tectable levels and was stimulated to proliferate into an epizootic by the recent warming trend. Increased parasite abundance in the new locations has then augmented its spread within the new range. If this hypothesis is correct, will a return to a more typical temperature regime, especially cold winters, push the parasite from its new range? In assessing this potential, it must be recalled that P. marinus has not only been detected in a new geographic range, but both prevalence and infection intensity are high in many areas. Because oysters arc abundant in these areas (e.g., Delaware Bay seed beds. Oyster Bay, Long Island Sound, and Cotuit Har- ! PRE-EPIZOOTIC EPIZOOTIC /TV POST-EPIZOOTIC Perception threshold TIME Figure 8. Schematic showing phases of an epizootic, including the pre-epizootic period when the existence of a pathogen is unknown. Adapted from Brown (1987). Pf:RK/NSUS MARINUS RaNGE EXTENSION 55 bor). the parasite population is also very high. Even if adverse conditions arc severe enough to destroy all but a very small frac- tion of the parasites, those surviving can be numerically abundant enough to rekindle an epizootic when conditions become more favorable. A recently developed mathematical population model of oyster-P. manniis interactions predicts that conditions, includ- ing low temperature, needed to terminate an epizootic are more extreme than those required to trigger one (Powell et al. 1996). Similarly. Burreson and Ragone Calvo (1996) concluded that a single unusually warm winter in Chesapeake Bay had a greater impact on the proliferation and spread of P. mannus than a cold winter did m its elimination. In fact, the cold winter of 1993-94 failed to end the epizootic in the northeast and winter temperatures in 1994—95 were again above average. Stopping or materially slowing the epizootic will require not one but a series of consec- utive cold (i.e., average or below average temperatures) winters and cool springs that will delay the summer infection buildup until late in the season. A restricted period during which lethal infec- tions can develop, kill oysters, and further disseminate infective particles appears to have helped end the I95()s epizootic in Dela- ware Bay. In estuaries where salinity is potentialh limiting, a number of consecutive years of above average rainfall at the same time would further help, as there appears to be a synergistic effect of low salinity and low temperature that is harmful to parasites (Burreson and Ragone Calvo 1993. Burreson et al. 1994, Chu et al. 1994, Ragone Calvo and Burreson 1994). In addition, substan- tial freshwater inflow into any water body should diminish the abundance of infective particles by flushing (Mackin 1956). The hypothesis that the recent wanning trend in the northeast- em United States, particularly the above average winter tempera- tures, has fostered the range extension of P. marinus needs to be tested further by examining trends in water temperatures and de- termining how closely actual seasonal temperature cycles (as op- posed to long-term trends) now associated with P mannus in the northeast approximate those in more southern locations that have supported the parasite for decades. In addition, continued efforts must be made to determine whether the parasite has become, or is becoming, adapted to lower temperatures. The development of lower temperature-adapted parasites might be more feasible now that large populations are established in a region where selection pressure exists for this trait. That P. marinus can adapt to low salinity has been demonstrated in vitro using cultured parasites (O'Farrell et al. 1995). Although it has not been demonstrated that a genetic change was involved, such adaptation could theoretically occur in nature and in response to low temperature as well as low salinity. Obtainine a better understanding of the causes for the ranee extension has both fundamental and practical implications. Re- searchers have long advised against the transportation of oysters from areas where a contagious disease agent is enzootic into "dis- ease-free" water (Rosenfield and Kem 1979, Ford 1992). The history of P. marinus demonstrates how risky it is to consider either areas or hosts as free of a pathogen merely because it has no obvious effect on the host or it cannot be otherwise detected. How restrictive should regulations on transportation of molluscs by commercial growers be if uncontrollable factors such as climate change are demonstrated to play a significant role in the spread of disease? This is an especially relevant question because the many other potential means of introduction occur despite such regula- tions. Now that P. marinus is found in New England, should restrictions be relaxed on introductions of southern oysters or other molluscs that might be carrying the parasite? What potential does transplantation between regions known to be enzootic for a patho- gen have to exacerbate or prolong a disease problem (see Bushek and Allen 1996)? The range extension by P . marinus should be a stimulus for a new discussion of reasonable and rational controls on the movements of oysters in commercial culture. ACKNOWLEDGMENTS I am indebted to many individuals who contributed to this paper. Harold Haskin established and maintained the long-term programs that documented the presence of P. marinus in Delaware Bay before 1990. Greta Christensen and Donald Kunkle per- formed all diagnoses between 1955 and 1963 and drafted an un- published report containing data used in this paper. After 1990, Jesselyn Gandy, Bob Barber, and Marty Chintala, helped by nu- merous student assistants, were responsible for diagnoses. Joe Dobarro and Royce Reed of the New Jersey Bureau of Shellfish- eries and Hill Bloom, Steve Fleetwood, and Larry Hickman of Bivalve Packing Co. provided boats for sample collection in Del- aware Bay. Walt Canzonier obtained information from oyster planters on shucking southern oysters. Numerous growers in the northeastern United States provided oyster samples for diagnosis. Thanks to Dave Bushek, Walt Canzonier, and John Kraeuter for helpful critique and discussion on early versions of the manuscript. Delaware Bay data were obtained through long-term funding from the New Jersey Department of Environmental Protection and its predecessors. Data from New England were obtained with funding from the Northeastern Regional Aquaculture Center under USDA Grant #90-38500-521 1 and from NOAA under the Oyster Disease Research Program, Grants #NA47FL0153 and NA57FL0042. This is Contribution No. 95-28 from the Institute of Marine Sci- ence at Rutgers and New Jersey Agricultural Experiment Station Publication No. D-32502-1-95. Andrews, J. D. 1955. Notes on fungus parasites of bivalve mollusks in Chesapeake Bay. Proc. Natl. Shellfish. Assoc. 45:157-163. Andrews, J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsiis muriniis and its effects on the oyster industry. Amer. Fish. Soc. Spec. Pi/bl. 18:47-63. Andrews, J. D. & W. G. Hewatt. 1957. Oyster mortality studies in Vir- ginia II. The fungus disease caused by Dermocystidium maniiiiin in oysters of Chesapeake Bay. Ecol. Monogr. 27:1-26. Bauersfeld. W. R., E. W. Moshinsky & C. E. Gumey. 1991. Water Re- sources Data. New Jersey, Water Year 1991 . U.S. Geological Survey. Bauersfeld. W R . E W. Moshinsky & C. E. Gumey. 1993. Water Re- sources Data. New Jersey, Water Year 1993. U.S. Geological Survey, LITERATURE CITED Bauersfeld, W. R E. W. Moshinsky & E. A. Pustay. 1992. Water Re- sources Data. New Jersey, Water Year 1992. U.S. Geological Survey. Brown, G. C. 1987. Modeling. In: J. R. Fuxa & Y. Tanada (eds). Epi- zootiology of Insect Diseases. John Wiley and Sons, New York, pp. 43-68. Burreson, E. M. & L. M. Ragone Calvo. 1993. The effect of winter temperature and spring salinity on Perkinsus marinus prevalence and intensity: a laboratory experiment. (Abstract) J. Shellfish Res. 12(1): 125. Burreson, E. M, & L. M. Ragone Calvo. 1996. Epizootiology of Perkinsus marinus disease of oysters in Chesapeake Bay. with empha- sis on data since 1985. J. Shellfish Res. 15:17-34. Burreson, E M , L. M. Ragone Calvo, J. F. La Peyre, F. Counts & K. T. 56 Ford Paynter. 1994. Acute osmotic tolerance of cultured cells of the oyster pathogen Perkinsiis mariims (Apicomplexa: Perkinsida). Comp. Bio- chenu Physiol. l09A:575-582. Bushek. D. 1994. Dermo Disease in the American Oyster. Genetics of Host-Parasite Interactions. PhD Dissenation. Rutgers University. New Brunswick, New Jersey. 189 pp. Bushek, D. & S. K. Allen. Jr. 1996. Races of Perkinsiis mwiinis. J. Shellfish Res. 15:103-107. Bushek, D., S. E. Ford & S. K. Allen. Jr. 1994. Evaluation of methods using Ray's fluid thioglycollate medium for diagnosis of Perkinsus marimis infection in the eastern oyster, Crassostrea virginica. Annu. Rev. FishDis. 4:210-217. Choi, K.-S,, E. A. Wilson. D. H. Lewis. E, N. Powell & S. M. Ray. 1989. The energetic cost of Perkinsiis mariniis parasitism in oysters. Quantification of the thioglycollate method. J. Shellfish Res. 8:117- 125. Chu, F.-L. E. & J. F. La Peyre. 1993. Perkinsiis nuiriniis susceptibility and defense-related activities in eastern oysters. Crassostrea virginica: temperature effects. Dis. Aquat. Org. 16:223-234. Chu. F.-L. E., J. F. La Peyre & C. S. Burreson. 1993. Perkinsiis marinii.\ susceptibility and defense-related activities in eastern oysters. Cras- sostrea virginica: salinity effects, J. Invertebr. Pathol. 62:226-232. Chu. F.-L. E.. A. K. Volety & G. Constatin. 1994. Synergistic effects of temperature and salinity on the response of oysters {Crassostrea vir- ginica) 10 the pathogen Perkinsiis manniis. (Abstract) J. Shellfish Res. 13(1):293. Cooper, K. R. & W Durfee. 1982. A one-year disease survey of four commercially important Rhode Island bivalve mollusks: Mya arenaria. Argopecten irradians irradians. Crassostrea virginica. and Merce- naria mercenaria. (Abstract) J. Shellfish Res. 2:93. Gushing, D. H. & R. R. Dickson. 1976. The biological response in the sea to climatic changes. Adv. Mar. Biol. 14:1-122. Ford, S. E. 1992. Avoiding the spread of disease in commercial culture of molluscs, with special reference to Perkinsii.s marimis (Dermo I and Haplosporidium nelsoni (MSX). J. Shellfish Res. 1 1(2):539-546, Ford, S. E. 1996. History and present status of shellfisheries from Bame- gat Bay to Delaware Bay. NOAA Technical Reports. (In press). Ford, S, E. & H. H. Haskin. 1982. History and epizootiology of Haplo- sporidium nelsoni (MSX). an oyster pathogen, in Delaware Bay, 1957-1980. J. Invertebr. Pathol. 40:118-141. Haskin. H. H. & J. D. Andrews. 1988. Uncertainties and speculations about the life cycle of the eastern oyster pathogen Haplospondiiiin nelsoni (MSX). Amer. Fish. Soc. Spec. Piibl. 18:5-22. Haskin. H. H. & S. E. Ford. 1979. Development of resistance to Min- chinia nelsoni (MSX) mortality in laboratory-reared and native oyster stocks in Delaware Bay. Mar. Fish. Rev. 41( l-2):54-63. Haskin. H. H.,L. A.Stauber&J. A. Mackin. \'^6b. Minchinia nelsoni n. sp. (Haplosporida, Haplosporidiidae): causative agent of the Delaware Bay oyster epizootic. Science 153:1414—1416. Haven, D. S.. W. J. Hargis & P. C. Kendall. 1978. The Oyster Industry of Virginia: Its Status. Problems and Promise. Virginia Institute of Marine Science, Gloucester Point, Virginia. 1024 pp. Hofmann. E. E.. E. N. Powell. J. M. Klinck & G. Saunders. 1995, Mod- elling diseased oyster populations I. Modelling Perkinsiis marimis in- fections in oysters. J. Shellfish Res 14( 1 ): 121-151 , Ingersoll, E. 1881. The Oyster Industry. Department of the Intenor, Wash- ington, DC. 251 pp. Jeffries. H, P. & W. C. I, Johnson, 1976. Petroleum, temperature, and toxicants: Examples of suspected responses by plankton and benthos on the continental shelf. In: B. Manowitz (ed). Effects of Energy-Related Activities on the Atlantic Continental Shelf. Brookhaven National Lab- oratory, pp. 96-108. Kochiss, J. M. 1974. Oystering from New York to Boston. Wesleyan University Press, Middletown, Connecticut, 251 pp. Mackin, J. G. 1956. Dermocystidiiirn marimiin and salinity. Proc. Natl. Shellfish. Assoc. 46:116-133. Mackin, J. G. 1962. Oyster diseases caused by Dermocyslidiiim marinum and other microorganisms in Louisiana. In: J. G. Mackin and S. H. Hopkins (eds). Studies on Oysters in Relation to the Oil Industry. Vol. 7. Publication of the Institute of Manne Science. Texas A & M Uni- versity, pp. 132-229. Mackin. J. G. & S. H. Hopkins. 1962, Studies on oyster mortality in relation to natural environments and to oil fields in Louisiana, In: J. G. Mackin and S, H, Hopkins (eds). Studies on Oysters in Relation to the Oil Industry, Vol. 7. Publication of the Institute of Marine Science, Texas A & M University, pp. 1-131, Mackin, J G . H M. Owen & A Collier. 1950 Preliminary note on the occurrence of a new protistan parasite, Dermocyslidiiim marinum n. sp. in Crassostrea virginica (Gmelin). Science 111:328-329. Mackin. J. G. & A. K. Sparks. 1962. A study of the effect on oysters of crude oil loss from a wild well. In: J. G. Mackin and S. H. Hopkins (eds). Studies on Oysters in Relation to the Oil Industry. Vol. 7. Publication of the Institute of Marine Science, Texas A & M Univer- sity, pp. 230-261. Myers, T. R. 1981. Endemic diseases of cultured shellfish of Long Island, New York: adult and juvenile American oysters (Crassostrea virgin- ica) and hard clams (Mercenaria mercenaria). Aquacidtitre 22:305- 330. Newman. M. W. I97I. A parasite and disease survey of Connecticut oysters. Proc. Natl. Shellfish. Assoc. 61:59-63. OFarrell. C. L.. J. F. La Peyre & E. M. Burreson. 1995. Acute osmotic tolerance of cultured cells of the oyster pathogen Perkinsus marimis acclimated to low salinity. (Abstract) J. Shellfish Res. 14(1):274. Otto. S. V. & G, E, Krantz. 1977, An epizootic of "Dermo'" disease in oysters in the Maryland portion of the Chesapeake Bay, (Abstract) Proc. Natl. Shellfish. Assoc. 67:121. Powell. E. N.. J. D, Gauthier. E. A. Wilson. A. Nelson, R, R. Fay & J. M, Brooks, 1992. Oyster disease and climate change. Are yearly changes in Perkinsus marimis parasitism in oysters (Crassostrea vir- ginica) controlled by climatic cycles in the Gulf of Mexico'? P. S.Z.N. I: Mar. Ecol. 13(31:243-270. Powell, E. N., E. E. Hofmann & J. M Khnck 1996. Modeling diseased oyster populations II. Triggering mechanisms for Perkinsus marinus epizootics, J. Shellfish Res. Quick, J. A., Jr. 1977. Fungus disease of oysters. In: C. J. Sindemiann (ed). Disease Diagnosis and Control in North American Manne Aqua- culture. Elsevier Science Publishers. Amsterdam, pp. 198-203. Ragone Calvo. L. & E. M. Burreson. 1994. Characterization of overwin- tering infections of Perkinsus marinus (Apicomplexa) in Chesapeake Bay oysters. J. Shellfish Res. 13(1 ):123-130. Ray. S. M. 1954. Biological Studies of Dermocystidium marinum. a Fun- gus Parasite of Oysters. Rice Institute. Houston. Texas. 1 14 pp. Rosenfield. A. & F. G. Kern. 1979. Molluscan imports and the potential for introduction of disease organisms. In: R. Mann (ed). Exotic Species in Mariculture. MIT Press, Cambridge, Massachusetts, pp. 165-191. Saunau, P.-G. 1991. Spread of Cyclope neritea (MoUusca: Gastropoda) along the north-eastern Atlantic coasts in relation to oyster culture and to climatic tluctuations. Mar. Biol. 109:299-.TO9, Sindermann. C, J. 1970. Principal Diseases of Marine Fish and Shellfish. Academic Press. New York. 369 pp. Soniat, T. M. & J. D. Gauthier. 1989. The prevalance and intensity of Perkinsus marinus from the mid-northern Gulf of Mexico, with com- ments on the relationship of the oyster parasite to temperature and salinity. Tiilane Stud. Zool. Sot. 27:21-27, Southward. A. J.. G, T, Boalch & L. Maddock. 1988, Fluctuations in the herring and pilchard fisheries of Devon and Cornwall linked to changes in climate since the 16th century, J. Mar. Biiil. Assoc. U.K. 68:423- 445. Southward. A. J.. E. I, Butler & L, Pennycuick, 1975, Recent cyclic changes in climate and in abundance of marine life. Nature 253:714- 717. Journal of Shellfish Research. Vol. 15. No. 1. 57-66, I9')6. LABORATORY INVESTIGATIONS OF SUSCEPTIBILITY, INFECTIVITY, AND TRANSMISSION OF PERKINSUS MARINUS IN OYSTERS FU-LIN E. CHU Virginia Institute of Marine Science School of Marine Science College of William and Mary Gloucester Point, Virginia 23062 ABSTRACT The protozoan parasite, Perkinsiis niarmus (Demio). has caused significant mortality in the eastern oyster, Crassostrea virginica. along the east coast of the United States and the Gulf of Mexico, since the 1950s. Because of its current expanded distribution and increased abundance, P. marinus is now considered more prevalent in the mid-Atlantic waters and the Chesapeake Bay in particular, than another protozoan pathogen, Haplospuridium nelsoni (MSX). The susceptibility, infectivity/palhogenicity, and transmission of P. marinus in eastern oysters were investigated in numerous laboratory studies. The intluence of environmental factors such as temperature, salinity, and pollution on the interaction between the host oyster and the parasite were also examined. Three P. marinus life stages, the meront, prezoosporangia, and bitlagellated zoospore, were found effective in transmitting the disease. The meront stage was more effective than the prezoosporangia stage in transmitting the disease in eastern oysters, suggesting that the meront is the primary transmission agent in nature. A dose of 10-10" freshly isolated P marinus cells oyster ' was required to cause infection by direct shell cavity injection. P. marinus susceptibility and disease progression were positively correlated with temperature, salinity, and number of infective cells the oyster encountered. Temperature appeared to be the most important factor, followed by the infective cell dose, and then salinity in determining the subsequent disease development in oysters. There was no significant interaction between temperature, salinity, and infective cell dose on the prevalence of disease in oysters. However, the interaction between either temperature and salinity or between temperature and P. marinus dose significantly intensified the disease. The Pacific oyster, Crassoslrea gigas. was less susceptible, but not completely resistant, to P. marinus compared to the eastern oyster, C. virginica. However, the Pacific oyster was intolerant of high temperature ( > 1 5°C) and low salinity ( < 10 ppt), thus vulnerable to high mortality under high temperature and low salinity environmental conditions. Pollution has the potential to enhance P. marinus susceptibility and infection in oysters . KEY WORDS: Pcrkinsus marinus. eastern oyster. Crassostrea virginica, disease, susceptibility, infectivity, transmission INTRODUCTION For the last 40 years, the protozoan parasite Perkin.^ii.', marinus (Dermo) has continuously caused severe mortality in the eastern oyster. Crassostrea virginica, from the Delaware Bay throughout the mid- Atlantic and Gulf coasts in the United States (Andrews 1988. Andrews and Ray 1988). To verify field observations and to achieve a better understanding of disease processes and transmis- sion dynamics in nature, numerous studies have investigated the susceptibility, infectivity. and transmission of this parasite since its discovery in oysters of Louisiana coastal waters (Mackin et al. 1950). The intluence of environmental factors on the host-parasite interaction has also drawn much attention. However, extensive laboratory studies on these subjects did not occur until the niid- 1980s. This chapter reviews and discusses laboratory studies on susceptibility, infectivity. disease processes, and transmission of P. nuiriims in eastern oysters. TRANSMISSION OF P. M.ARINUS IN OYSTERS Although the life cycle of P. marinus is still not completely known, three life stages, meront. prezoosporangia. and biflagel- lated zoospores, have been identified and described (Perkins 1966, Perkins 1988). Immature meronts (merozoites) usually found in the phagosomes of hemocytes are 2-4 jxm and coccoid. Meronts (10-20 (i,m) are mature merozoites with an eccentric vacuole that often contains a refringent vacuoplast. The mature meront. an 8 to 32 cell stage enclosed within a mother cell wall, is a sporangium (schizont. 10—4-0 ^JLnl). When meronts are placed in fluid thio- glycoUate medium (FTM) for 4—5 days, they develop into pre- zoosporangia (hypnospores). which are sometimes observed in moribund and dead oyster tissues and can enlarge to 150 (xm. Prezoosporangia are characterized by having a large vacuole and an eccentric nucleus adjacent to the cell wall. After incubating thioglycoUate-cultured prezoosporangia in sea water for 4—5 days, zoosporulation (production of bitlagellated zoospores) usually oc- curs. However, it is unclear whether prezoosporangia released in sea water from moribund and deceased oysters would zoosporulate in nature. Laboratory study of P. luariiuis disease and transmission pro- cesses in oysters began in the early 1950s when the thioglycollate tissue assay became available (Ray 1952). The infectivity and pathogenicity of different life stages, the required dosages for infection, and routes of transmission were of great interest. Infectivity and Pathogenicity of Different Life Stages The disease caused by P. marinus is infectious and can be transmitted from infected to uninfected oysters. Placing uninfected oysters in the same container with infected oysters, the uninfected oysters ultimately become infected (Ray and Mackin 1954). All three identified life stages, meront, prezoosporangia and biflagel- late zoospore, can cause P. marinus infection in oysters. How- ever, it is not known which life stage is most effective and the principal stage for transmitting disease in the field. The infectivity and pathogenicity of the two life stages, meront and prezoospo- rangia, were compared in our laboratory (Volety and Chu 1994). We inoculated lO'* meronts or prezoosporangia into the shell cav- ity of individual oysters, at 2I-25°C and 14—20 ppt, that were collected from an area outside the normal geographic range of P. marinus (Damariscotta River, ME) and measured infection prev- alence and intensity in these oysters over a range of time intervals. 58 Chu Infections developed within 40 days in oysters inoculated with either meronts or prezoosporangia. Infection prevalence (Fig. 1 A) and intensity (Fig. 2A) increased with time in both groups. How- ever, prevalence of infection was significantly higher in oysters inoculated with meronts than with prezoosporangia. At 75 days postchallenge, intensities of infections ranged from light to mod- erate in oysters exposed to meronts. whereas only light infections were noted in oysters exposed to prezoosporangia. A similar trend was demonstrated in oysters collected from the Ross Rock area of the Rappahannock River, VA, an area that typically has low prev- alence of P. marinus infection (Fig. IB. Fig. 2B), with the ex- ception that infection first appeared in oysters 15 days after expo- sure to prezoosporangia. However, this may represent an infection acquired in the field since the organisms for this experiment were collected from a location affected by P. marinus. Infection was not detected in the Maine oysters sampled at 20 days after exposure to prezoosporangia or meronts. It is uncertain whether the different infection rates and inten- sities (Figs. 1 and 2) in oysters between these two experiments were caused by batch variation in infectivity of meronts and pre- zoosporangia or were due to difference in response to P. marinus between the two oyster populations. Previous studies demon- strated that P . marinus infection rates varied among oyster strains or populations (Burreson 1991 . Chu and La Peyre 1993a). The P. marinus susceptibility of six oyster strains, including two native strains from various disease enzootic areas in the southern Ches- apeake Bay. one native strain from the Delaware Bay, and a Hap- losporidium nelsoni (MSX)-resistant strain, were compared by Burreson (1991). All the native strains were found to have lower disease-caused mortality than the MSX-resistant strain. The native 100 so t 60 o J .^ 40 c 20 a ^ □ Prezoosporaroa ■ Meronts D Control 20. 40. 60. 65. 75. 100 40 20 1 I ES Prezoosporan(>a ■ Meronts D Control 15. 25. 40, 65. hAjTiber of days (post chaltenge) Figure 1. (A) C. virginica. P. marinus prevalence in oysters (Dama- riscotta River, ME) at 20, 40, 50, 65, and 75 days postchallenge by meronts or prezoosporangia (N = 8-9 oysters per sampling date). (B) P. marinus prevalence in oysters (Rappahannock River, VA) at 15, 25, 40, and 65 days postchallenge by meronts or prezoosporangia (N = 5-8 oysters per sampling date). MERONTS SPORANGIA TREATMENTS MERONTS SPORANGIA TREATMENTS Figure 2. C. virginica. Intensity of P. marinus infection in oysters in- oculated Kith meronts or prezoosporangia after 75 days (A, oysters from Damariscotta River, ME) and after 65 days (B. oysters from Rappahan- nock River, VA). strains continued to grow, whereas the MSX-resistant strain did not. after being infected by P. marinus. Similarly. Chu and La Peyre (1993a) noted that the progression and development of P. marinus infection differed, to some extent, among three oyster populations from the Chesapeake Bay. Results from our experiments indicate that meronts and pre- zoosporangia can effectively transmit P. marinus disease among oysters. Meronts are probably the primary transmission agent in nature. Oysters inoculated with this life stage resulted in higher infection prevalence and intensity than oysters inoculated with prezoosporangia. The proliferation rate of P. marinus meronts (and merozoites) cultured at 5. 12. 20. and 28°C was positively correlated with temperature (Volety 1995). The high prevalence of infection in oysters inoculated with meronts can be interpreted as a result of rapid multiplication of meronts at warm temperatures (i.e.. 21-27°C). Moreover, it has been reported that 99% of P. marinus cells found in the water of the upper Chesapeake Bay in the warmer months (March to October) between 1992 and 1993 resembled the meront stage (Dungan and Roberson 1993). The lower infection rate of oysters inoculated with prezoosporangia compared to meronts is puzzling. The fate of inoculated prezoo- sporangia within the oyster is not known, although division of Susceptibility, Infectivity, and Transmission 59 prezoosporangia into meront-like structures by schizogony has been observed in Hi viiro cultures (La Peyre 1993, Perkins, per- sonal communication). It is possible that there is a lag time for the inoculated prezoosporangia to partition to meront stage or that culture of meronts in FTM affects the infectivity of the subsequent prezoosporangia. In in vitro culture, prezoosporangia develop into zoosporangia in sea water, zoosporulate. and subsequently release biflagellated zoospores (Perkins 1966. Chu and Greene 1989). However, the production of zoospores by meronts or prezoospo- rangia in oysters or in cells isolated from oyster tissue without FTM treatment has not been documented. The infectivity and pathogenicity of biflagellated zoospores are unknown, partly because there are still missing links in the P. marinus life cycle. For example, the biflagellated zoospore, which has not been seen in oyster tissue, originates from the zoospo- rangium that develops in sea water outside of the host. How it enters the host and how it develops into a meront stage once it penetrates into the host tissue are unknown. In the past, production of zoospores regularly occurred from prezoosporangia cultured in sea water (Perkins 1966, Chu and Greene 1989). and we routinely infected oysters with biflagellated zoospores. In experiments conducted in the early 1980s, direct injection of a dose of 10^ zoo- spores per oyster into the shell cavity successfully induced infec- tion in oysters (Chu, unpublished data; Morris Roberts, Virginia Institute of Marme Science, unpublished results). Nevertheless, recent attempts in several laboratories to culture prezoosporangia to zoosporulation have not been successful. Occasionally, some prezoosporangia zoosporulate but produce only a few biflagellate zoospores. Thus we have been unable to include the zoospore stage in infectivity and pathogenicity comparisons. The biflagellated zoospore can swim and has an apical com- plex, presumably for host entry by penetration (Perkins 1988). If zoosporulation occurs in nature after prezoosporangia are released from moribund or deceased oysters, then the role of biflagellated zoospore m transmission oi P . marinus should not be disregarded or underestimated. Unlike the meront stage, which is passive and therefore must rely on water currents of flushing rate to be trans- mitted between oysters, the biflagellated zoospores are motile. Moreover, the usual winter temperatures of subtropical climates would not be able to kill the prezoosporangia embedded in the oyster tissues. Prezoosporangia can withstand temperatures as low as 4°C, and zoosporulation has been observed in prezoosporangia, previously incubated at 4°C, after they were transferred to 28°C (Chu and Greene 1989). Minimal Dose for Infection The number oi P . marinus cells required to transmit the disease from mfected, dying, and gapmg oysters to uninfected oysters has been in question for a long time. To determine whether the oyster mortality caused by P . marinus is correlated to the infective cell concentrations, Mackin ( 1962) inoculated oysters with minced tis- sues containing various concentrations ( 10-10'') of meronts which had been incubated in FTM for 24 hr. He found that a dose of 1 .0 X 10" to 5.0 X 10" cells was required to cause mortality within 41 days and mortality rate was positively correlated with the inocu- lated infective cell numbers. However, Mackin only monitored the oyster mortality caused by exposure to the P . marinus infective cells but not the infection rate or disease intensity in oysters. In a study to compare the P . marinus susceptibility in four different laboratory-reared stocks of C. virginica, Valiulus (1973) observed infection in experimental oysters at 105 days after they received a dose of 10 cells per oyster injected into their shell cavity. Recently, Chu and Volety (unpublished observation) tested the responses of eastern oysters to different doses of meronts and prezoosporangia using oysters from Damariscotta River, ME. In two experiments, a dose-dependent response of P. marinus infec- tion was found in oysters exposed to 0. 10, 10", lO'*, and 10'^ meronts or prezoosporangia at 25°C for 8-12 weeks (56-84 days). In both experiments, infection prevalence was found to signifi- cantly increase with increasing dose of P. marinus infective cells inoculated into the shell cavity of the oysters (Fig. 3A, data from one of the two experiments). A similar pattern was shown in infection intensity, expressed as weighted incidence (WI = sum of disease intensity rating/total number of oysters examined. Fig. 3B). Again, the meront stage caused much higher P. marinus infection prevalence and intensity in oysters than prezoosporangia. The lowest dose that initiated a P. marinus infection was between 10 and 10" meronts or prezoosporangia per oyster. No mortality occurred during these studies. The dose response in oysters to in vitro cultured P. marinus was studied by Bushek ( 1994). In three separate experiments, he challenged oysters with different doses of cultured infective cells via direct injection into shell cavity or adductor muscle or by feeding. He found that oysters fed a dose as high as 10^ cells oyster"' did not succumb to infection, whereas light infections 10 100 10C00 100000 Number of P. marinus cells exposed B M«rontt I I Prezooiporongla 2 1.8 1.6 • 1.4 I 1^ c « 5 0.6 0.4 0.2 0 10 100 10000 100000 Number of P. marinus coils exposed ^^1 Meroi^ls I PrezooBporangIa Figure 3. C. virginica. Prevalence (A) and intensity (B) of P. marinus in oysters (Damariscotta River, MEl 56 days after inoculation with 0, 10, 10", lO'', or 10' meronts or prezoosporangia (N = 14-15). 60 Chu developed within 50 days with doses above 10'* cells oyster"' when inoculated into the shell cavity or the adductor muscle. In the latter two cases, infection intensity increased as dose increased. Similarly, infection developed in oysters in 28 days after two sequential inoculations of laboratory-cultured P. marinus cells (1.1 X 10^^ and 3.3 x 10'' cells oyster"') into the shell cavity, while no infection developed in oysters at 56 days after feeding three doses of laboratory-cultured P. marinus (e.g., 3.56 x U) . 3.36 X 10^ and 1.36 x 10^ cells oyster"') (Perkins 1994). Laboratory-cultured P. marinus cells may be less infective/ pathogenic than those freshly isolated from infected oyster tissues. Generally, oysters infected with laboratory-cultured meronts (and merozoites) did not exhibit an infection rate as high as the oysters infected with meronts freshly isolated from oyster tissues (Volety and Chu unpublished results, Perkins unpublished results). Dosing oysters with 10" cultured meront cells per oyster. La Peyre et al. ( 1993) detected only light infections in 8 weeks, and no mortality was noted. Recently, Chintala et al. (1995) compared the infec- tivity of laboratory-cultured P. marinus cells and P. marinus cells isolated from infected oysters. They challenged oysters with the same number of natural isolates or cultured P marinus and fol- lowed mortality. At 12 weeks postchallenge, 15% of the oysters challenged with natural isolates died with heavy P. marinus in- fection while only 7.5% of the oysters challenged by laboratory- cultured P. marinus died. In contrast to the above findings, Gau- thier and Vasta (1993) found heavy infections in oysters 4—5 weeks after two biweekly inoculations of 2 x \0'' cultured meront cells and mortality occurred at 6-8 weeks. The cultured meronts used in Gauthier and Vasta's work seem more virulent than either Bushek's (1994) or La Peyre's (1993). It is not known whether the difference in infectivity in cultured meronts is due to the dif- ference in composition of culture media or the origin of the meront isolates. The medium developed by Gauthier and Vasta (1993) is quite different from the one used by Bushek ( 1994), La Peyre et al, (1993), and Chintala et al. (1995). The latter three used meronts cultured in medium developed by La Peyre et al. (1993). Bushek ( 1994) also tested the infectivity of isolates of P. mari- nus from different locales. Infectivity or virulence varies between isolates of P. marinus. Significantly heavier infections were found in oysters challenged by two Atlantic Coast isolates (Chesapeake and Delaware Bays) than in oysters challenged by the two Gulf Coast isolates (Barataria Bay. LA, and South Bay Laguna Madre, TX). The findings of Bushek's study ( 1994) suggest the existence of geographic P. marinus races. Exactly how P. marinus infective cells, discharged from car- riers and/or released from infected gapers, enter a new host is unclear. Entry through filtration/feeding is assumed to be the main route of P. marinus transmission in nature, so it is of interest that shell cavity injection is more effective than feeding in infection induction (Ray 1954, Bushek 1994, Perkins 1994). The infective cells, either meront or prezoosporangia, are particularly sticky (personal observation). They would easily adhere to the mantle and gills while passing through the host's shell cavity during the feeding and/or filtration process. If this is taken into consideration, infection generated through shell cavity inoculation of infective cells may be comparable to the natural condition. It was demonstrated through the dose response experiments that a dose of 10-I0~ freshly isolated P. marinus cells oyster" ' is required to cause infection by direct shell cavity injection (Chu and Volety, unpublished observation). Considering the abundance of P. marinus (3,000-19,000 1" ' water) found in the water during warmer months of the year (Dungan and Roberson 1993), and the filtration rate (>8 1 hr~ ', Galtsoff 1964) of oysters in nature, it may be obvious how P. marinus infection is transmitted. As Bushek (1994) speculated, chronic feeding of high levels of P. marinus may be required to cause infection. However, once 10- 10' infective cells are trapped in the shell cavity, infection will result (Valiulus 1973. Chu and Volety unpublished observation). SUSCEPTIBILITY AM) INFECTIVITY; TEMPERATURE AND SALINITY EFFECTS It has been known, since the epizootics caused by P. marinus in oysters in the 1950s, that the distribution and abundance of the parasite in the field are limited by temperature and salinity (see reviews by Andrews 1988, Andrews and Ray 1988). Beginning in the 1950s, studies have been carried out to determine the relation- ship between P. marinus incidence in oysters and temperature and salinity under laboratory-controlled conditions. Hewatt and An- drews (1956) studied the effect of high and low temperatures on P. marinus infection in oysters by holding oysters in the laboratory at 15°C and oysters in trays in the York River at 26-28°C for 6 weeks after feeding the oysters minced, heavily infected oyster tissues. They found that low temperature substantially reduced the oyster mortality caused by the parasite, thus suggesting that low temper- ature may inhibit the parasite's activity. In an earlier study, they (Andrews and Hewatt 1957) also found that oysters maintained in an aquarium at 15°C did not develop infection after 6 weeks, while oysters maintained at room temperature (20-2I°C) were all in- fected by the parasite after 2 weeks. In two experiments, Ray ( 1954) compared the mfection rate of artificially infected oysters maintained at two salinity ranges (26-28 and 10-14 ppt; 24—29 and 10-15 ppt). He found that the developmental rate of infection in the oysters at the low salinity range was delayed. From these early studies, it was concluded that temperature and salinity affect the infection rate and development of the parasite in oysters. However, the lowest test salinities in Ray's study (1954) were 10-15 ppt and 10-14 ppt, which are much higher than the lowest salinities (3-6 ppt) that P. nuirinus can tolerate (Chu and Greene 1989), and only two temperatures or salinities were com- pared forP. marinus infection in these studies. Thus no correlation can be drawn between P. marinus incidence and temperature or salinity using these data. Moreover, in these investigations, oys- ters developed infection and died within 2 weeks, indicating that the experimental oysters may have been previously infected in the field. The relationship, documented in field studies, between P. marinus infection in oysters and temperature and salinity, was reaffirmed through more detailed and comprehensive laboratory studies by Chu and her associates (Chu and La Peyre 1993b, Chu et al. 1993a). In two separate studies, they examined the effects of temperature and salinity on P. marinus susceptibility and infection in oysters after challenging individual oysters with lO'' freshly isolated meronts. Their results showed that P. marinus suscepti- bility and infection in oysters were significantly correlated with experimental temperatures and salinities. Disease prevalences and infection intensity decreased as temperature decreased (Fig. 4). Likewise, P. marinus prevalence and intensity (WI) in meront- challenged oysters were positively related to salinity (Fig. 5). Oys- ters at 3 ppt exhibited lower P. marinus prevalence and intensities than those at 10 and 20 ppt. Heavy infections were found only in oysters at 10 and 20 ppt. However, in the salinity experiment, the Susceptibility, Infectivity, and Transmission 61 £ 60 10C 10PC 15C 1BPC 20C 20PC 25C 25PC Temperature ( C) \ 3C 3PC 10C 10PC Salinity (ppt) 20C 20PC 10C 10PC 15C 15PC 20C 20PC Temperature ( C) 25C 25PC Figure 4. C. virginica. P. marinus infection prevalence (A) and inten- sity (B) in oysters (Rappahannock River, VA) at 10, 15, 20, and 25°C (N = 40). C = Control; PC = oysters challenged by P. marinus. unchallenged oysters showed a pattern of P. marinus infection siinilar to the challenged oysters. This is believed to be an expres- sion of latent infection established in the field (a subsample (N = 40] taken in this experiment at the time of collection exhibited 12.5% light infection). Therefore, salinity as low as 3 ppt did not eliminate infection but did prevent intensification. Similar results were noted by holding oysters infected with P . marinus in water of 6 ppt (Ragone and Burreson 1993). TRANSMISSION ECOLOGY Field observations point to temperature and salinity as two important environmental factors regulating the activity oi P. mari- nus. Certainly, the dosage of P. marinus infective cells is also critical. Therefore, it is important to examine the impact of the interaction among these three crucial factors on the outcome of the disease process. Recently, Chu et al. (1994) investigated the re- sponse of oysters challenged by two different doses (2.5 x 10^ or 2.5 X 10'' freshly isolated meronts oyster^ ') off. marinus at nine salinity-temperature combinations: i.e., 10, 15, and 25°C at 3, 10, and 20 ppt. Results agree with their previous findings on the susceptibility and infectivity and dose response studies. Increased infection prevalence and intensity (Fig. 6) occurred at high tern- Figure 5, C. virginica. P. marinus prevalence (A) and intensity (B) in oysters (James River, VA) al 3. 10, and 20 ppt (N = 19-24). C = control; PC = oysters challenged by P. marinus. peratures and salinities and there was a dose-dependent response to infective particles. When the effects of temperature, salinity, and infective cell doses and their interaction on P. marinus suscepti- bility and infection intensity were analyzed using logistic regres- sion and a log linear model, temperature was found to be the most important factor influencing the susceptibility to P. marinus and subsequent disease development in oysters. This was followed, respectively, by the dose of infective particles and salinity. Sim- ilarly. Fisher et al. ( 1992) found temperature to be more influential than salinity on P. marinus intensity and mortality in oysters col- lected from the Gulf of Mexico and maintained at different test temperatures (I8-27°C) and salinities (6-36 ppt). It has also been reported that fluctuations in incidence of the parasite in oysters in the Gulf of Mexico region are more sensitive to changes in tem- perature than to salinity fluctuations (Mackin cited in Ray 1954). Applying these results to the field observations may explain why P. marinus infection in oysters is dominant in the summer months at mid-Atlantic waters and is generally confined to subtropical regions. It is also true that, in nature, river water inputs and/or fresh water runoff not only dilute the salinity, but reduce the in situ concentration of P. marinus infective cells in estuaries, thus pro- tecting oysters to some extent from infection (Mackin 1962). 62 Chu 20ppt C Dt D2 C D1 D2 C D1 D2 Tfeatmenis 3 PPT 10 PPT 20 PPT Figure 6. C. virginica. P. marinus prevalence (A) and intensity (B) in oysters (Damariscotta River, MF.I at different temperature and salin- ity regiines after challenge with two different doses of P. marinus infective cells. C = control: Dl = 2.5 x lo' meronts oyster'': D2 = 2.5 X 10'' oyster"'. N = 7-15 per group at all temperature-salinity treatments, except the groups at 25°C and 3 ppt treatment (N = 3—1 per group). However, along the Gulf Coast, unlike the Chesapeake Bay. water usually stays warm year-round. Therefore, the seasonal shift of P. marinus infection is more likely to be associated with the change of salinity caused by rainfall and river runoff rather than temperature. A program to monitor the regional distribution and yearly trends of P. iiuiriiuis prevalence and intensity in relation to climate changes in the Gulf of Mexico between 1986 and 1989 indicated that the yearly shift of P. marinus prevalence was related to the rainfall/river runoff (salinity) rather than temperature (Pow- ell et al. 1992). The effect of three-way interactions among temperature, salin- ity, and nieront doses on disease prevalence was found to be insignificant in our study (Chu and Volety. unpublished observa- tion). There was a significant two-way interaction, however, be- tween temperature and salinity, and between temperature and meront dose, on intensity of P. marinus infection. Based on data calculated from field samples, an effect of temperature and salinity interaction was noted for P. marinus infection intensity (weighted prevalence) in two different field studies conducted on the Gulf Coast (Soniat 1985. Soniat and Gauthier 1989). A COMPARISON OF CRASSOSTREA GIGAS AND C. VIRGIMCA: EFFECTS OF TEMPERATURE AND SALINITY ON SUSCEPTIBILITY TO P. MARINUS The eastern oyster. C. virginica, has historically supported a major fishery on the east coast of the United States. Because of severe mortality, beginning from the late 1950s, in oyster popu- lations caused by P. marinus and H. nelsoni in the mid-Atlantic region, introduction of a non-native species, the Pacific oyster iCrassostrea gigas). to the waters of this region was proposed to revitalize the oyster fishery. The Pacific oyster has been success- fully introduced and cultured along the west coast of the United States and in Europe. This oyster species is rarely infected by the protozoan parasite, Bonamia ostreae. that has caused severe losses of the European oyster (Ostrea edulis) industry in Europe and on the west coast of the United States over the last decade (Grizel 1985. Elston et al. 1987. Grizel et al. 1988). Results of recent studies using outdoor flow-though sea water systems with quar- antined effluent also indicate that the Pacific oyster is less suscep- tible than the eastern oyster to P. marinus (Meyers et al. 1991. Barber and Mann 1994). Pacific oysters usually thrive in waters of salinity higher than 18 ppt and temperature =£15°C. though they may be able to tol- erate a temperature as high as 35°C and salinity as low as 10 ppt (see review by Mann et al. 1991). Information regarding temper- ature-salinity tolerance in C. gigas is, however, limited and the definitive temperature and salinity tolerance of this species has not been tested in the laboratory. The mid- Atlantic climate is rela- tively warm, between temperate and subtropical, and the ecosys- tem of the Chesapeake Bay is complex. The salinity range of oyster habitats in the Chesapeake Bay varies seasonally and ranges from 0 to >20 ppt. The water temperature of most tributaries of the Chesapeake Bay can reach 28-29°C and persist for more than 2 months during the summer. The oyster pathogen, P. marinus. can persist in salinities lower than 5 ppt and the epizootics caused by this parasite are elevated at high temperatures and salinities (Andrews 1988, Burreson and Andrews 1988). Therefore, the competence of the Pacific oysters against P. marinus under dif- ferent salinity and temperature regimes is of particular concern. Chu et al. (1993b) evaluated, in the laboratory, the suscepti- bility of diploid (2N) and triploid (3N, estimated to be 95%) C. gigas to P. marinus compared with C. virginica at three test tem- peratures (10, 15, and 25°C) at a salinity of 20-22 ppt and at three test salinities (3, 10 and 20 ppt) at a temperature of 19-22°C. Their findings (Fig. 7) show that, generally. Pacific oysters were more tolerant to P. marinus infection than eastern oysters at the tem- perature and salinity regimes tested. However, at IO°C. P. mari- «((i-challenged 3N C. gigas had prevalences higher than P. mari- /!i«-challenged 2N C. gigas and C. virginica. In C. virginica, moderate and heavy infections developed in both control and P. »i(/n/ii/i-challenged groups at 25°C and in P. mar/HHi-challenged groups at 20 ppt. No heavy infection was found in C. gigas in any treatment. Moderate infections were detected only in diploid C. gigas at 10 and 25°C. Since much higher infection prevalences were found in the control (unchallenged) C. virginica. at any given temperature and salinity treatment, than in unchallenged 2N and 3N C. gigas (Fig. 7). the authors believed that part of the recorded prevalence and intensity in C. virginica was attributed to the ex- pression of hidden infection carried over from the field. Unfortu- nately, the techniques, thioglycollate tissue (Ray 1952) and he- molymph (Gauthier and Fisher 1990) assays, currently employed Susceptibility, Infectivity, and Transmission 63 10 16 26 Temperature ( C) 3 CanUo* (N-M-41) C.GIGASI2NI C.GIGASONI The investigation by Chu et al. (1993b) also revealed that Pa- cific oysters are much less tolerant than eastern oysters to low salinity and high temperature. Heavy non-P. »i(in>!Mi-related mor- tality occurred during their study, in both diploid and triploid Pacific oysters at salinities 20 ppt and below and temperatures higher than I5°C (Tables 1 and 2). High non-disease-relatcd mor- tality (70%) was also recorded in Pacific oysters in conjunction with low salinity (<20 ppt), in a study carried out to compare the growth and mortality of C. fiigas and C. vir'^inicii challenged with P. marinus (Barber and Mann 1994). These results suggest that C . gigiis may survive the P . muiinus challenge, but may die under the environmental conditions prevailing m the mid-Atlantic, particu- larly the southern Chesapeake Bay. Moreover, the shells of this species held in water of lower Chesapeake Bay (i.e., York River, VA) were found to be quite susceptible to invasion by the poly- chaete Polydara sp. (Burreson and Mann 1994). ^Hl ~^B^ ^B "Hm Sallnlly (ppl) ^^ CONTROL (N = 26-40) ^| CHALL (N = 23-41) C VIRGINICA C.GIGAS(2NI C, GIGAS I3N1 Figure 7. C. virginica and C. gigas, P. marinus pre>alence in C. vir- ginica (Rappahannock River, \ A) and C. gigas at 10, 15, and 25°C (A, N = 35-41) and at 10 and 20 ppt (B, N = 23-t3l. for routine P. nuiriiuis diagnosis apparently were not sensitive enough to detect cryptic infection. Therefore, in the future, eastern oysters from an area free of P. marinus should be employed for this kind of study. RELATIONSHIP BETWEEN POLLUTION AND P. MARINUS SUSCEPTIBILITY IN OYSTERS As described earlier, susceptibility and advancement of P. marinus infection in oysters are associated with environmental temperature and salinity. However, it is uncertain to what extent the increased distribution and intensification of P. marinus infec- tion in nature arc attributable to environmental degradation. Pol- lution has been hypothesized to contribute to some aquatic epi- zootics, although this link has not been adequately examined. To evaluate this hypothesis. Chu and Hale (1994) investigated the effects of a complex mixture of sediment pollutants on the sus- ceptibility of oysters to P . marinus. They exposed oysters to dif- ferent dilutions (i.e.. 0. 15. 30%) of water-soluble fractions (WSF) generated from sediments collected from Elizabeth River, a heavily polluted sub-estuary of the Chesapeake Bay. The Eliz- abeth River sediments used for the study were grossly contami- nated with polycyclic aromatic hydrocarbons, characteristic of cre- osote. The WSFs generated from the sediments were dominated by lower molecular weight aromatic hydrocarbons and heterocyclic compounds (Table 3). The mean concentration of aromatic pol- lutants in the WSFs (representing more than 100 compounds) was 4.08 mg r ' (S.D. = ±0.399). These oysters were then chal- lenged with P. marinus meronts. Pollutant exposure in the labo- ratory was found to enhance preexisting P marinus infections in TABLE 1. Mortality of C. virginica and C. gigas during temperature acclimation and after challenge with P. marinus Parameter Mortality ( # of deaths) during acclimation Mortality (# of deaths) after P. marinus exposure Total mortality C/r) during experiment** C. virginica C. gigas (2n) C. gigas (3n)* 10 15 25 10 15 25 10 15 25 Temperature (°C) (N = 80) (N = 80) (N = 80) (N = 79) (N = 81) (N = 85) (N = 82) (N = 82) (N = 111) 1.3 10 17.5 1.3 111 21 32.9 2,4 4.9 32 29.7 An estimated 95'7c of oysters were triploid. % = # of dead oysters/initial total number of oysters. 64 Chu TABLE 2. Mortality of C. virginica and C. gigas during salinity acclimation and after challenge with P. marinus C. virginica C. gigas (2n) C. gigas (3n)* Parameter salinity (ppt) 3 10 20 3 10 20 3 10 20 (N = 79) (N = 84) (N = 81) (N = 77) (N = 86) (N = 95) (N = 52) (N = 78) (N = 81) MortaUty (# of deaths) during acclimation Mortahty ( # of deaths) after f". marinus exposure Total mortality (%) during experiment** 44 37 20 30.2 16.8 12 17 37.1 10 17 33.3 An estimated 957f of oysters were triploid % = # of dead oysters/initial total number of oysters. oysters from an area affected by P. marinus and to increase the susceptibility to experimental infection in oysters from an area outside the normal geographic range of P. marinus (Fig. 8). Both occurred in a dose-dependent manner. Recently. Fisher et al. (1995) and Anderson et al. (1995) tested, independently, the effects of tributyltin (TBT) on P. mari- nus progression in eastern oysters. Both research groups observed that exposure to TBT significantly elevated mortality caused by P. marinus in oysters, when compared to non-TBT-exposed oysters. TBT exposure enhanced the disease prevalence (Anderson et al. 1995) and increased slightly the progression of P. marinus infec- tion in oysters (Anderson et al. 1995, Fisher et al. 1995). Pollutants may reduce disease resistance by causing physiolog- ical stress in the host or suppressing certain host immune mecha- nisms (Anderson 1996). Alternatively, pollution may affect dis- ease susceptibility by elevating infection pressure through an in- crease in the number and activity of infectious organisms. It is currently unknown whether the increased P. marinus infection observed in pollutant-exposed oysters is attributable to heightened P. marinus virulence or decreased host resistance, although the latter is suspected. Results from the above studies do suggest that environmental degradation may increase the epizootic, although P. marinus-cdused disease is known to be predominantly exacer- bated by elevated temperature and salinity. A previous study by Winstead and Couch (1988) noted that P. marinus expression TABLE 3. Concentrations (S.D.), in mg I"', of the major organic contaminants, detected in representative water-soluble fractions (WSF), generated from Elizabeth River sediments and control water (filtered York River water) (N = 3) Analyte Control WSF Naphthalene <0.001 1.510(0.520) Acenaphthene <0.001 0.424 (0.033) 2-Methylnaphthalene <0.001 0.224(0.018) Phenanthrene <0.001 0.210(0.026) Fluorene <0.00l 0.201 (0.031) Dibenzofuran <0.00l 0.195(0.012) 1 -Methylnaphthalene <0.001 0.151 (0.022) Carbazole <0.00l 0.148 (0,013) e 15-' N=33 N.33 Percent of WSF PC 50- / / / / / / / PC ^ 46- H~ 40- 35- m "'^"~^^""™ 30- ■ 25- m ""^-^"^^ 20- PC 15i ^m* 101 H 5 PO PO P PO ■'; 0 ^ ■rn^A / /^Z7/ /^=7/^ / N=20 N = 20 IS Percent o( WSF Figure 8. C. virginica. P. marinus prevalence in oysters after exposure to 0, 15, and 30% dilutions of WSF generated from contaminated estuarine sediments. PO = nonchallenged oysters; PC = P. marinus- challenged oysters. (A) Oysters (N = 23-33) from Rappahannock River, VA, were exposed to WSF dilutions for 35 days and then were challenged with P. marinus. Postchallenge WSF exposure was then continued for an additional 21 days. (B) Oysters (N = 20) from Dam- ariscotta River, ME, were exposed to WSF dilutions for 35 days and then were challenged with P. marinus. Postchallenge WSF exposure was then continued for an additional 35 days. Susceptibility, Infectivity, and Transmission 65 appeared at uncharacteristically low temperatures after toxicant exposure. CONCLUSION AND FUTURE STUDIES All three identified P . nuirinn.s life stages, meront. prezoospo- rangia, and bitlagcllated zoospore, are effective in transmitting disease. The meront stage is more effective compared to prezoo- sporangia stage in creating P. marinus infection in eastern oysters and thus may be the primary transmission agent in nature (al- though the mfectivity and pathogenicity of the bitlagcllated zoo- spore stage were not included in the comparison). Results of lab- oratory studies are consistent with field observations and clearly demonstrated that P. marinus susceptibility and disease advance- ment are positively correlated with temperature, salinity, and the number of infective cells to which the oyster is exposed. Among the above three environmental factors, temperature is the most important factor followed by the infective cell dose controlling the susceptibility to P. marinus and subsequent disease development in oysters. Salinity was a less influential factor than temperature and infective cell concentration. There was no significant three- way interaction among temperature, salinity, and infective cell dose on the prevalence of disease in oysters in the laboratory, but the two-way positive interactions between either temperature and salinity or between temperature and P marinus dose significantly intensified the disease in oysters. Culture of oysters in tributaries with high river water input and/or fresh water runoff may effec- tively dilute P. marinus infective elements, thus providing some level of protection to oysters from infection. The Pacific oyster. C. gigas. is less susceptible to P. marinus than the eastern oyster. C. virginiia. However, they may not survive if introduced into Ches- apeake Bay tributaries, since they arc unable to adapt well to low salinity and high temperature environmental conditions. Pollution has the potential to enhance P. marinus susceptibility and infection in oysters. To further understand the transmission and disease processes of P. marinus in eastern oysters, the life history of this parasite needs to be completely determined, it is not known why laboratory-cultured meronts seldom reach zoosporangia stage and why prezoosporangia derived from FTM-cultured meronts no longer zoosporulate. The infectivity and pathogenicity of the bi- tlagellated zoospores life stage need further examination. Further- more, we know little about the fate off. marinus cells after they enter the host. Finally, sensitive techniques are needed to detect P. marinus at low infection intensities without sacrificing the oyster. ACKNOWLEDGMENTS The author would like to thank Dr. Aswani Volety for his invaluable help in graphic works and Drs. Robert Hale. Ken Webb and Morris Roberts for their critical review s of the first draft of the manuscript. Virginia Institute of Marine Science Contribution no. 1955. LITERATURE CITED Anderson. R. S, 19%. Interactions of Perkinsiis marinus with humoral factors and heniocytes of Crassoslreii virf;iiuca. J. Shellfish Re\. 15:127-134. Anderson, R. S.. M. A. Unger. & E. M. Burreson. 1995. Enhancement of Perkinsus marinus disease progression in TBT-exposed oysters (Crassostrea vir^inica). Eighth International Symposium "Pollutant Responses in Marine Organisms." April 2-5. 1995. Monterey. Cali- fornia. Andrews. J.D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marinus and its effects on the oyster industry. Amer. Fish. Soc- Spec. Publ. 18:47-63. Andrews, J. D.. & W. G. Hewatt. 1957. Oyster mortality studies in Vir- ginia. II. The fungus disease caused by Dermocystidium marinum in oysters of the Chesapeake Bay. Ecol. Monogr. 27:1-25. Andrews, J. D. & S. M. Ray. 1988. Management strategies to control the disease caused by Perkinsus marinus. Amer. Fish. Soc. Spec Piihl. 18:206-224. Barber, B. J. & R. Mann 1994, Growth and mortality of eastern oysters. Crassostrea virginica (Gmelin 1791 ). and Pacific oysters. Crassoslrea gigas (Thunberg. 1793) under challenge from the parasite. Perkinsus marinus. J Shellfish Res. 13:109-1 14. Burreson. E. M. 1991. Status of the major oyster diseases in Virginia — 1990. A summary of the annual monitoring program. Marine Resource Report 91-1. Virginia Institute of Manne Science. Gloucester Point. Virginia. Burreson, E. M. & J. D, Andrews. 1988. Unusual Intensification of Ches- apeake Bay oyster diseases during recent drought conditions. Proc. Occawi 88:799-802. Burreson, E. M. & R. Mann. 1994. Field exposure of triploid Crassoslrea gigas to Haplospondium nelsoni (MS.Xl and Perkinsus marinus (Dermo) in the lower Chesapeake Bay. J. Shellfish Res. 13:293. Bushek, D. 1994. Dermo disease in American oysters: Genetics of host- parasite interactions. Ph.D. Dissertation. Rutgers, The State University of New Jersey, New Brunswick, New Jersey, pp. 1-189. Chintala, M. M., K. A. Alcox. S. E. Ford, & D. Bushek. 1995. Cultured Perkinsus marinus cells: A possible mechanism for virulence differ- ences. Milford Aquaculture Seminar. Mllford. Connecticut. Chu. F.-L. E. & K. H, Greene. 1989. Effect of temperature and salinity on in vitro culture of the oyster pathogen. Perkinsus marinus (Api- complexa: Perkinsea). J. Inveriebr. Pathol. 53:260-268. Chu. F.-L. E. &J. F. La Peyre. 1993a. Development of disease caused by the parasite. Perkinsus marinus and defense-related hemolymph factors in three populations of oysters from Chesapeake Bay. USA. J. Shell- fish. Res. 12:21-27. Chu, F.-L. E. & J. F. La Peyre. 1993b. Perkinsus marinus susceptibility and defense related activities in eastern oysters Crassostrea virginica: Temperature effects. Dis. Aquat. Org. 16:223-234. Chu, F. L. E. & R. C. Hale. 1994. Relationship between pollution and susceptibility to disease in the eastern oyster. Crassostrea virginica. Mar. Environ. Res. 38:243-256. Chu. F.-L. E. J. F. La Peyre. & C. S. Burreson. 1993a. Perkinsus mari- nus infection and potential defense-related activities in eastem oysters. Crassoslrea virginica: Salinity effects. J. Invertehr. Pathol. 62:226- 232, Chu. F.-L, E.. C. S, Burreson. A. Volety. & G, Conslantin. 1993b. Per- kinsus marinus susceptibility in eastem {Crassoslrea virginica) and Pacific (Crassostrea gigas) oysters: temperature and salinity effects. J. Shellfish Res. 12:127. Chu, F.-L. E.. A. K. Volety. & G. Conslantin. 1994. Synergetic effects of temperature and salinity on the response of oysters {Crassostrea virginica) to the pathogen. Perkinsus marinus. J. Shellfish Res. 13: 293. Dungan. C. F. & B. S. Roberson. 1993. Flow cytometnc quantification and analysis of Perkinsus marinus cells present in estuarine waters. Final Report. NCAA NMFS Oyster Disease Research Program. Con- tnbution No. NA16FL0406-OI . Elston. R.. M. Kent. & M. Wilkinson. 1987. Resistance o{ Oslrea edtilis to Bonamia osirea infection. Aquaculture 64:237-242. Fisher. W. S.. J. D. Gauthier. & J. T. Winstead. 1992. Infection intensity of Perkinsus marinus disease in Crassostrea virginica (Gmelin. 1791 ) from the Gulf of Mexico maintained under different laboratory condi- tions. J. Shellfish Res. 1 1:363-369. Fisher. W. S., L. M. Oliver, E. B. Sutton, C. S, Manning. & W. Walker. 1995. Exposure of eastem oysters to tributyltin enhances the seventy of 66 Chu PerkinsKs mariniis ai^-SgB>»g^^^ Figure 16. Mature trophozoite in whicii a vacuoplast (asterisk) has heen formed, pnsuniahly from inlravacuolar \acuoplast material (V). I'niike this eell, in most preparations the vacuoplast and vacuoplast material have the same electron density . Lipoid droplet (L); intranuclear virus-like particles (arrowheadl. TKM. 26,0(IOx. Figure 17. Eight-cell tomont. C'entriole with electron-dense inclusion (Cl; (lolgi bod) (G); developing eccentric vacuole (Va); mitochondrion (Mt). Note regular pattern of cytokinesis yielding cuneiform, immature trophozoites. TEM. 28,000x. 74 Perkins ^^^^^^HjB^k • V \ I^^Ti i*" •^ |Pf^ ''^MKV ,_^\ ^^"^^^-s/^^^X jW" -'.^V \ '' Wfl"^ 1^*^' ^ 9^^ ^^^yd ,■1®. ^-^^v • •*, ..ea*-" Figures 18 and 19. Hypnospores found in adductor muscle of previously uninfected ( . virginica which was exposed to zoospores of Perkinsus sp. from M. halthica. The hypnospores were induced to form in FTM and are indistinguishable from those of/*, inarinus. Note the large eccentric vacuole iVal in each cell around which there is a thin layer of cytoplasm pressed against the cell wall (W). Nucleus (arrowhead); lipid droplets (arrow). BF. Fig. 18 = ISSx; Fig. 19 = SOOx. Figure 20. Hypnospore of P. inarinus equivalent to those in Figs. 18 and 19. Note the extreme size of the eccentric vacuole, the thick cell wall and the large expanses of membrane-free cytoplasm with ribosomes. Nucleus (N); lipid droplet (L). TEM. l(),600x. Structure of Pi-:HKii\si's makisus 75 most immature and mature trophozoites enlarge to various sizes. often into the 3()-80 (jim range. I have observed extremes of up to 480 (xm in diameter. In the proecss. the vacuoplast disappears, the cell wall becomes thicker and the eccentric vacuole enlarges mark- edly resulting in a thin layer of cytoplasm pressed against the cell wall (Figs. 18-20 and 35). The mitochondria may disappear or are reduced to thin profiles with few cristae probably as a result of being held in the anaerobic or microaerophilic conditions extant in the medium. Otherwise, the ultrastructure is the same as in the mature trophozoites except that the nucleus enlarges and contains a prominent nucleolus. These cells are termed hypnospores (see terminology section in the DISCUSSION). In his original observation of enlargement in FTM. Ray ( 1452) also found that a diluted ( 1:25) aqueous solution of Lugol's Kl-I, stained the hypnospores blue, blue-black or blue-green. Smce the solution could be applied to pieces of oysters after holding in FTM and then squashed under a coverslip. a rapid technique for detec- tion of the pathogen was developed. Based on their studies in which thc> used cytochemical and cn/ymc extraction techniques. Stein and Mackin ( 1457) suggested that the thick hypnosporc wall consists of hemicellulose and cellulose. They speculated that the staining reaction in LugoPs iodine solution in the absence of acid hydrolysis is due to the combination of the two polysaccharides, a combination w hich is lacking in the trophozoites. The latter do not stain blue, blue-green or blue-black in Lugol's solution. Perkins and Menzel (1%7) demonstrated a 2()-3()-|j.m diameter, fibrillar network in hypnosporc walls which had been isolated and treated with a 50'* solution of 0.25N NaOH in conmiercial bleach (Clo- rox). This network most probably represents the polysaccharide complex. If the hypnospores arc washed free of the FTM and placed in estuarine or sea water of ca. 20-30 ppt. zoosporulation by pal- intomy ("palintomic zoosporulation" shortened herein to '"zoo- sporulation") may occur. However, as opposed to what has been observed over the last 20 years, in recent years preparations of P. marinus hypnospores show a low incidence of zoosporulation without release of the zoospores (see comments in the Materials and Methods section). Therefore, rather than using earlier pub- lished micrographs, zoosporangial development is illustrated using cells of Perkiiisiis sp. from M. hallhica, where the process is indistinguishable from that seen in P imiiimis (see Fig. lA-L in Perkins and Menzel 1966). The morphogenesis consists of diges- tion or subdivision of cytoplasmic inclusions and lipoidal droplets so that the cytoplasm becomes finely granular with active Brown- ian movement. A lens-shaped clear area (rarely two areas) under the cell wall can be seen to form and the nucleus enlarges mark- edly (Figs. 21 and 22). Over the clear area a discharge pore is formed around which the edges of the pore are ragged and out- wardly flared, indicating that pore formation is an explosive event Coupled with pore formation is the extension of a discharge tube (sometimes two) which unfolds from the lens (Fig. 23). At the base of the tube and occluding the pore is an ovoid plug. Ultra- structural observations of the pore complex reveal that the plug is an enlargement of a secondary cell wall layer which is synthesized beneath the preexisting wall. The tube is in continuity with the plug (Perkins and Menzel 1467). As in palintomy observed in host tissues, the eccentric vacuole is subdivided into small vacuoles distributed randomly in the cy- toplasm (Fig. 24) as the cell contracts and pulls away from the cell wall. Presumably much of the fluid originally in the eccentric vacuole is squeezed into the space between the cell and its wall. The details of mitosis have not been elucidated; however, fine structure of the lew mitotic profiles which have been observed indicate that the nuclear envelope may remain intact and the cen- trioles serve as the spindle pole bodies. It appears that deep, chan- nel-like invaginations of the nuclear envelope are formed with the nuclear envelope intact. Bundles of microtubules are found in the invaginations attached to the envelope at kinetochore-like struc- tures. The details are being elucidated and will be described in a separate publication. Thus, mitosis resembles that seen in some dinotlagellates such as Amphidiniuin carterae (Dodge 1987). However, the dinokaryotic condition of coiling of the chromatin is not observed in P . marinus. The chromatin appears as electron- dense granular aggregates of varying density whether in interphase or during mitosis. After the first karyokinesis. cytokinesis occurs by an equatorial pinching in half of the binucleate cell to yield two separate and distinct cells (Fig. 25). This process is repeated so that after each cycle, each nucleus and whole cell become smaller and smaller until. |ust before the final cytokinetic event, the cells are no longer spheroidal but have become shaped like a kernel of nee (Figs. 26 and 27). At this time synthesis of the tlagella has occurred, as c\ idcnced by movement of the cells w ithin the zoosporangial wall and by the fact that free quadritlagellatcd zoospores can sometimes be observed instead of the normal biflagellatcd ones, presumably as a result of premature release from the zoosporangium and not isogametic copulation. A final cycle of karyokinesis, followed by cytokinesis, then occurs, and bitlagellated zoospores can be seen to swim actively in the zoosporangium for periods of up to 2 days (Fig. 28) before the plug at the base of the discharge tube disin- tegrates and the zoospores swim out of the zoosporangium through the tube (Fig. 29). They measure 2-3 x 4—6 jim. are rounded at the anterior end and are pointed at the posterior end with a sub- apical indentation (groove?) on one side about one third of the cell body length from the anterior extremity. The flagella arise from the indentation (Figs. 30 and 31). Before the final cycle of karyokinesis and cytokinesis, synthe- sis of the apical complex, which characterizes the Apicomplexa. is initiated. The complex consists of a conoid, polar ring, rhoptries. microneme-like structures and subpellicular microtubules (Figs. 31 and 32). The conoid is anteriorly situated and composed of microtubular units arranged in a helical coil forming a truncated cone which is open along one side (Fig. 32). Around the anterior end is a ring of electron-dense material forming the anterior polar ring to which are attached microtubules of the cytoskeleton which extend beneath the plasmalemma to the posterior end of the zoo- spore where there is a posterior ring. The microtubules do not appear to attach to the latter ring (Perkins 1976a). Attached to and below the anterior polar ring and surrounding the anterior third of the conoid is a truncated cone open at either end consisting of what appears to be the same material which forms the ring (Perkins 1976a). Flattened alveoli (vesicles) are found between the plas- malemma and the cytoskeleton at the anterior end of the cell but are not repeated posteriorly as in other Apicomplexa. A thin layer of granular material is present beneath the plasmalemma. instead of alveoli, starting at the posterior margin of the alveoli and ex- tending beneath the plasmalemma around the rest of the cell. The rhoptries are membrane-bound, tlask-shaped sacs of elec- tron-dense material which have the neck of the tlask inserted into the lumen of the conoid and extend almost to the anterior end of the cell. A group of long membrane-bound organelles are also found with the anterior ends inserted into the conoid lumen and 76 Perkins Figures 21-29. /oii^porulation in Perkinsm sp. from .\t. balthica after placement of the cells in cstuarine water from KTM. The morphogenesis is indistinguishable from that of P. marinus. After 42 hr, 45 min in estuarine water at 25°C (Fig. 21 1 the nucleus iNi has enlarged markedly and there is a thickening of the cytoplasm (arrow) away from the nuclear region (N). A plaque of wall material then formed beneath the original cell wall (Fig. 22; arrow), followed by formation of a pore in the cell wall and a discharge tube (D) from the plaque (Fig. 23). At the base of the pore is a plug of secondary wall material also derived from the plaque (Fig. 23: eccentric \acuole, \ ; nucleus. N). The cell then subdivided, the eccentric vacuole forming a frothy cytoplasm in which the first karyokinesis occurred (Fig. 24: edge of pore, arrow, beneath which is the plug, PI). First cytokinesis occurred (Fig. 25) follovied by more cycles of karyokinesis and cytokinesis (Figs. 26-28) until hundreds of motile zoospores had been formed. After the plug disappeared and the zoospores had matured, they swam through the discharge tube to the outside. Free zoospore (Z); zoospore emerging from the zoosporangium (arrow). BF and PC. 440x. extending almost the full length of the cell. They have been called rectilinear micronemes as a term of convenience by Perkins (1976a) only because they have a diameter much the same as micronemes, are membrane bound and contain electron-dense ma- terial as in other Ap[comple\a. They differ in that the length (s much greater, and they are not completely filled with electron- dense material. Another type of organelle is attached to the pos- teriorly extended part of the conoid in a row of membrane-bound, long and wavy units of about the same diameter as the rectilinear micronemes and contains electron-dense mater(al. These extend the length of the cell and have been termed conoid-assoe[ated micronemes, also as a term of convenience. It is not known wheth- er either type of "micronemes" is homologous to the micronemes of the Apicomplexa (Figs. 31 and 32). In the anterior end of the zoospore is found a U-shaped large vacuole (Fig. 31 ) which contains electron-dense material (not vis- ible in Fig. 31) resembling the droplets which contribute to the vacuoplast material found in mature and immature trophozoites. The material is in contact with the vacuole membrane and does not fill the vacuole. Other organelles in the zoospore (nclude a single Golg( body located next to the nucleus, one mitochondrion with tubular cr[stae located on the dorsal side of the cell ( if one accepts that the side to which the tlagella are attached is ventral) and a ventrally located nucleus with well-defined heterochromatin. Lipid droplets are present in the cell, generally in the posterior end, and are often visible as a single refractive inclusion when viewed using phase contrast microscopy. The posteriorly directed flagellum is almost straight and has no mastigonemes. The anteriorly directed nagellum is about three times as long as the posterior one and is always more or less coiled whether contracting and engaging in generation of forward motion of the cell (Fig. 30) or extending in the recovery mode. The coiling has no obvious, fixed pattern or form except that at the end of the recovery mode it generally is more tightly coiled than when it has completed contraction to generate forward motion. Filamentous mastigonemes are found along one side of the anterior flagellum in groups of about five, and at the base of each group is a distally pointed, flexed and spur-like unit (Fig. 33). Units resembling the filamentous mastigonemes are found in cytoplasmic vacuoles ot differentiating zoospores. Presumably, assembly of the mastigo- Strl'cture of Perkinsvs marimis 77 Figure 30. Actively swimming zoospore witii rounded anterior end of cell facing tlie top of tiie micrograph. Note coiling of anterior flagellum (arrowhead! around the cell hody in retraction phase of flagellar movement and the straight, posteriorly directed, posterior flagellum (arrov\ I. The subapical indentation in the right side of the cell body is the site from which both llagella arise. PC with strobe Hash. 2,2()0x. Figure 31. Longitudinal section through zoospore in v\hich can be seen the conoid (arrow), rhoptry (arrowhead), general region of flagellar attachment (*), alveoli (A), conoid-associated micronemes (CMl, Golgi body ((J), microtubules of cytoskelelon (Mc). rectilinear micronemes (Mi), dorsally located mitochondrion (Mt), nucleus (N), polar ring (Po), subpellicular granular layer (S) and vacuoles (Va). TEM. 40,fl00x. From Perkins (1987). Figure 32. Diagram of part of apical complex including conoid (arrow ), conoid-associated micronemes (CM), microtubules of cytoskelelon (Mc) and polar ring (Po). From Perkins (1976ai. nemes occurs in the vacuoles which are derived, (n turn, from the Golgi body. The kinetosome has a prominent electron-dense, cylindrical inclusion in the mid-region of its lumen w ith struts connecting it to the A and B microtubules of the kinetosome Inplct blades (Perkins 1988). The distal end of the kincto.some contains a cup-shaped structure with the open end of the cup facing the proximal end (Fig. 34). Several niicrotubular rootlets arise from a complex, electron-dense region at the proximal end of the kinetosome. The other details of the flagellar apparatus have not been elucidated. At the base of the tlagelluni is a transition plate where the central pair of microtubules of the axoneme terminate. Distal to 78 Perkins "■^*=«»^ ':iJ^ Figure 33. Uranyl acetate-stained whole mount of tlie mid-region of a zoospore'* anterior flagellum. Filamentous mastigonemes (arrow), spur (arrowhead). There is a single row of mastigonemes and spurs attached periodically with a cluster of four to five mastigonemes accompanying each spur. TEM. 130,000x. Figure 34. Kinetosome of zoospore (arrow ) and environs in which can be seen a cylindrical inclusion (I); cup-like structure (C); terminal helix (H); terminal plate of central pair of axoneme microtubules (T); and presumptive microtubular organizing centers (*) at base of kinetosome from which microtubular rootlets arise. TEM. 156,000 x. Figures 35—39. Hypnospores in which the progression of events leading to tuhuUir outgrowth torniation. followed by subdivision into unicells. can be seen. The cell wall around the tubular outgrowth is thinner than that of the original wall of the hypnospore. Nucleus (N); cytoplasm (arrowheadi; bulging of primary cell wall over the region of outgrowth formation (arrov\ I; unicells (I'). Figs. 35 and 36 = DK'; Figs. 37-39 = PC. Figs. 35 and 36 = 800x; Figs. 37 and 39 = 40(lx: Fig. 38 = 300x. Figure 40. Two hypnospores connected by a tubular outgrowth. The outgrowth (arrow I of the cell on the right has fused with the outgrowth (arrowheadi of the cell on the left. Whether cytoplasmic mixing and nuclear fusion occur in cells connected in this manner has not been determined. PC. 300 x. Figures 41—16. Immature and mature trophozoites and tomonts in axenic culture. Fig. 41 = mature trophozoite; Fig. 42 = presumptive binary Tission of mature trophozoite: Fig. 43 = first division of mature trophozoite engaged in palintomy. The eccentric vacuole has not subdivided as is seen in palintomy in infected oysters (see Fig. lOl (interface between two immature trophozoites = arrow). Figs. 44 and 45 = presumptive tomonts with four or five and more than eight immature trophozoites, respectively, in which the cells have large vacuoles and are irregular in shape; Fig. 46 = four mature trophozoites, in each of which is seen a single vacuoplast which appears attached to the eccentric vacuole membrane. Nucleus (N); vacuoplast (V): eccentric vacuole (Val. DIC. Figs. 41 and 46 = I.OOOx; Figs. 42 and 43 = 800 x; Fig. 44 = 900x; Fig. 45 = 750X. 80 Perkins the plate is a transitional helix wrapped around the central pair of microtubules (Fig. 34). Although it has not been visualized, presumably the zoospore makes contact with an epithelial cell of the host and the rhoptries induce a depression in the host cell surface so that the parasite is internalized in a vacuole as occurs with other Apicomplexa (Per- kins 1992). The flagella would be lost in the process as would the apical complex, and the cell would round up to become a cell resembling an immature trophozoite. Alternatively, the zoospores could be phagocytized by hemocytes with the loss of the same organelles. Nonzoosporulating Development of Hypnospores It has been observed by Ray (1952. 1954) that hypnospores while in FTM will form tubular outgrowths with bulbous termini, structures which he suggested were involved in microconidial de- velopment as in the Entomophthorales of the Eumycota. 1 have found in recent years that when hypnospores are removed from FTM and placed in sea water such outgrowths are formed, but I have not seen them while the cells are still in fTM. The out- growths begin as a thickening in the cytoplasm away from the nucleus. This thickening bulges the cell wall outward (Fig. 35). A distinct cytoplasmic protrusion then forms encased in cell wall material thinner than that of the preexisting hypnosporal cell wall (Figs. 36 and 37). This protrusion continues to grow outward until it is as much as two to three times as long as the diameter of the hypnospore (Fig. 38). The nucleus has been observed to remain within the spheroidal part of the cell wall after the outgrowth is quite long. At some undetermined time the nucleus probably en- ters the outgrowth and divides because small elongate cells are cleaved from the protoplasm in the tube. Presumably the cells are nucleated but this has not been demonstrated. The cells spill out of the distal end of the outgrowth and are found clustered around the outgrowth (Fig. 39). In other hypnospores, the protoplast may migrate out of the distal end of the outgrowth, form a bulbous mass and then subdivide to fonn similar-sized cells as above. This bulbous mass resembles the structures described by Ray (1952. 1954). An outgrowth may fuse with a neighboring cell (Fig. 40). but it is not clear whether mixing of cytoplasm or karyogamy occurs. In addition, an outgrowth may have irregularities in the surface which resemble limited branching of hyphae in the lower fungi. No ultrastructural investigations of the outgrowths have been con- ducted. Development in Axenic Culture When grown in the Kleinschuster and Swink ( 1993) medium, mature trophozoites (Figs. 41 and 46) of P. marimis were ob- served to have a wider range of sizes, from 3.8 to 43.2 |jLm in diameter (N = 100). than those observed in oyster tissue (range = 2.9-1 1 .6), and the mean was 11 .0 fim (S.D. = 9.4). The latter is about twice the diameter of mature trophozoites in the oyster where the means were 6.4 and 5.5 in fixed and unfixed cells, respectively. The differences could be due to the enriched nutri- tional environment of the culture medium. In addition, the vacu- oplasts (Fig. 46) were sometimes observed to be fixed to the eccentric vacuole membrane and did not oscillate by Brownian movement as in those from oyster tissue. In the ultrastructural observations of cultured mature trophozoites (La Peyre et al. 1993 and Fig. 15). the cellular structure resembled that observed in uncultured cells from oyster tissue (Fig. 14 and Perkins 1969) including the presence of ca. 50-nm-diameter virus-like particles in the nucleoplasm. The primary differences between cells cultured in the Klein- schuster and Swink medium and naturally occurring cells involve the forms which cytokinesis assumes as visualized at the light microscope level. No ultrastructural observations of cytokinesis were made. Both binary fission and palintomy appear to result in immature trophozoite formation in the cultured cells. In the former, the nucleus appears to divide, the cytoplasm migrates to form polar thickenings (Fig. 42). and then it is assumed that the cell pinches in half in the equatorial region without any other subdivision. Various degrees of equatorial pinching have been observed; however, completion of division has not been docu- mented. Asymmetric binary fission (budding) of a smaller cell from the parent meront also appears to occur (see also Fig. 6 of La Peyre et al. 1993). Palintomy in cultured cells occurs most often, as observed in infected oysters; however, there is a degree of plasticity or vari- ability of form which is not seen in oyster tissue. The eccentric vacuole often does not markedly subdivide prior to cytokinesis as seen in Fig. 10. but rather the vacuole may be halved with each cell division or it may lie at one pole of the cell and the cytoplasm divides at the other end (La Peyre et al. 1993; their Figs. 3-5). This suggests that progressive cleavage (multiple fission) is in- volved. In some cases there may be only two. three or four cells formed in a crescent of cytoplasm at one pole without any change in the vacuole. In Fig. 43. cytokinesis has just occurred or is nearing completion, and the interface between two immature tro- phozoites is visible. After the first cytokinesis the immature tro- phozoites may become irregular in size and shape (Figs. 44 and 45). the irregularity being more apparent the more numerous the daughter cells. However, most often palintomy occurs as seen in uncultured cells in the host. More observations are necessary to determine whether the infrequently observed "frothy" cells sim- ilar to the one in Fig. 45 are in fact viable or are anomalous and undergo lysis without completion of cytokinesis. DISCUSSION An attempt has been made herein to provide a useful reference source for recognizing P. mannus at both the light and electron microscope levels of detail, summaries of which can be found in Figs. 47 and 48. There are a number of heterotrophic, mostly saprobic. but some pathogenic species of coccoid protist in the estuarine and marine environments which could be confused with P. marinus when the cells are observed free of the host. C. vir- ginica. For example, this could occur in observing cultures of microorganisms or samples of sea water, sediments or tissues of organisms presumed to be parasitized. The most likely to present difficulties with identification are members of the family Thraus- tochytriaceae Sparrow 1943 of the phylum Labyrinthomorpha Page in Levine et al. 1980 (Corliss 1994). Members of this family superficially resemble Perkinsus spp. in cell size, shape and mode of cytokinesis, but generally lack the large eccentric vacuole and form bitTagellated zoospores with a bilateral array of tubular mastigonemes, ectoplasmic nets from specialized organelles termed sagenogens and laminated walls consisting of circular plates (Perkins 1974, 1976c; Porter 1990). They normally do not invade tissues of C. virgimca. but I have observed (unpublished data) small clusters oi Lahyrinlhiiloides-hke cells in the connective Structure of Phrk/xsus marim's 8i tissue beneath epithelia of the oyster. Labyrinlhiiloides halwiulis invades the host tissues and causes disease of juvenile abalone (Bower 1987): other species of Thraustochytriaceae cause diseases of other molluscs (Polglase 1980, Jones and O'Dor 1983. McLean and Poner 1987). In addition to the differences noted above which are ultrastruc- tural or require the observation of zoosporulation. one can differ- entiate between P . marinus and the thraustochytriaceous species by use of the Ray FTM technique. If (he cells enlarge in FTM. form thick cell walls and acquire a markedly large eccentric vac- uole, then they are most probably a species of Perkinsus. Further- more, if the application of LugoFs iodine solution as prescribed by Ray ( 1952) yields a blue, blue-black, black or blue-green staining of the cell wall after incubation in FTM, then the certainty of the identification is greater. No other protistan cell walls have been shown to stain these colors in Lugol's solution unless pretreated with strong acid, with the possible exception of some ciliate cysts (Perkins and Menzel 1967). The final confirmation would be to induce zoosporulation and find hiflagcllated zoospores with a uni- lateral row of filamentous mastigonemes and spurs on the anterior flagellum and an apical complex. However, as noted above, in- duction of zoosporulation in P. marimix has been shown to be difficult in recent years, as opposed to other species of Perkinsus. Short of ultraslruclural information, the best characters which can be used to differentiate between the thraustochytriaceous forms and P. marinus. in the light microscope and without treatment in FTM, are the presence, in mature trophozoites of P. marinus. of Figure 47, Developmental cycle of P. marinus in C. virginica. Imma- ture Iruphozulte 1 1) becomes a mature trophozoite (4), in the process acquiring a large eccentric vacuole with a vacuoplast. free in the vac- uole, and a nucleus uith a centrally located nucleolus. Palintomy (5-7) occurs during which the nucleus becomes progressively smaller through the first three karyokineses and the nucleolus becomes invis- ible in the light microscope. Most often S-16 immature trophozoites (range = ca. 4-641 are liberated from the tomont (5-7) through a tear in the wall (7 to 1). Figure 48. Zoosporulation of P. marinus in sea water, free of the host. Mature trophozoite enlarges markedly, losing the vacuoplast (1 to 2): a discharge tube and pore, occluded by a plug of secondary wall material, develop in the wall |3); palintomy results in the formation of numerous biflagellated zoospores (4 to 6) during which the nucleus and individual cells are reduced in size and the nucleolus becomes invisible in the light microscope; dissolution of the wall plug and lib- eration of the swimming zoospores then occur through the tube (6 to 7), Zoospores swim to or are pulled into an oyster's mantle cavity, ultimately resulting in new infections being established in or beneath the gill, mantle or gut epithelia. Presumably the zoospores lose their flagella and apical complex, become rounded and result in cells which can then be termed immature trophozoites. a large eccentric vacuole often with a vacuoplast and hyaline cy- toplasm with a few refringent inclusions. The cytoplasm of the thraustochytriaceous species is most often finely granular and a large eccentric vacuole is infrequently seen, never with a promi- nent vacuoplast. Dungan and Roberson (1993) significantly advanced the ability to microscopically detect cells of Perkinsus spp. through the use of antibody labelling. As they noted, this technique will undoubtedly prove to be very useful in detecting the presence of the pathogens in environmental samples and determining whether there are cells of Perkinsus sp. which are not detected in host tissues when his- tological or FTM techniques are used. Despite the high level of specificity involved in the use of antibody labelling, there will nevertheless be the necessity to compare the results with observa- tions of cytological and fine structure to eliminate any doubts. There still may be cellular stages in the life cycle yet to be detected and those techniques may prove to be critical in finding such stages, particularly in samples from the host-free environment (Li et al. 1994). Of particular interest may be the daughter cells formed from outgrowths of hypnospores (Figs. 35 to 39). The question arises as to whether the formation of these daughter cells represents the beginning of a saprobic phase in the life cycle or is a terminal phase which results in death of the cells if they do not enter C. virginica or another host organism. In addition to the immunolabeling approach to detecting cells of Perkinsus spp., the recent characterization of the small rRNA gene of P. nuirinus (Fong et al. 1993) has provided the ability to synthesize DNA probes which should be species specific. Such a capability should permit detection of P. marinus as opposed to any other microbe, including distinction of one species of Perkinsus from another. A wide diversity of bivalve molluscs are known to harbor cells of Perkinsus sp. or spp. as evidenced by the formation of zoosporangia in FTM (Ray 1954, Andrews 1955, Goggin and Lester 1987). The sizes of cells in Table 1 are about the same as recorded earlier (Mackin et al. 1950; Ray and Chandler 1955; Mackin 1962; Perkins and Menzel 1967; Perkins 1969, 1976b, 1988), with the exception of the upper limits for mature trophozoites where sizes of 10 |jim (Perkins 1976b), 20 pim (Perkins 1969, 1988) and 20 or, rarely, 30 |j.m (Ray and Chandler 1955) have been recorded. The largest mature trophozoite measured for this paper was 11.6 ^x.m (Table I). An explanation for the larger sizes that I recorded in earlier publications (and probably also those of Ray and Chandler 1955) IS the observation that in moribund oysters mature tropho- zoites may become much larger than usual and are probably cells equivalent to those formed in FTM (i.e., hypnospores) (Fig. 13; Perkins 1988). However, this is speculation because zoosporula- tion has never been demonstrated in any P. marinus cells removed from infected oyster tissue and placed directly in estuarine or sea water without FTM treatment. Zoosporulation can be induced in the larger cells of Perkinsus sp. from M . balthica by removing them from the host and placing them in estuarine water (Perkins 1968, Valiulis and Mackin 1969). If it is accepted that cells off. marinus larger than about 12 (xm in diameter in oyster tissue are actually hypnospores. then there is a natural equivalent to zoo- sporulation observed in the large cells (hypnospores) derived from FTM. Terminology As noted in the MATERIALS AND METHODS section, ter- minology used to denote stages in the life cycle of Perkinsus spp. 82 Perkins over the years has varied and has not been entirely satisfactory due in part to uncertainties concerning the phylogenetic affinities of the molluscan parasites and the unique cellular stages in the life cycle. The following justifications are presented for the use of terminol- ogy which is new to descriptions of Perkinsus spp. and for reten- tion of some previously used terms. Changes are being proposed due to the recent evidence for Perkinsus spp. having close affin- ities to both the Dinoflagellata and the Apicomplexa and due to the recognition that stricter use of the terms merogony. merozoites and meronts is justified. Palintomy is used herein because it denotes most accurately what occurs in Perkinsus spp., i.e.. a "rapid sequence of binary fissions, typically within a cyst and with little or no intervening growth, resulting in production of numerous, small offspring cells"" (Margulis et al. 1993). This is in keeping with Chatton (1937), who coined the term and considered palintomy to be di- vision where at each successive stage the cells became smaller and smaller: ". . . il se scinde coup sur coup en deux puis en quatre. et les scissions se poursuivent sans intervalle de croissance com- pensatrice. de sort que les produits sont beaucoup plus petits que Telement initial." Thus the term is the same as successive bipar- titioning (Perkins 1988). It has been used to denote asexual cel- lular proliferation in some parasitic dinotlagelletes (Cachon and Cachon 1987) and ciliates (Lynn and Small 1990. Margulis et al. 1993). Thus, for Perkinsus spp. palintomy is proposed as a replace- ment for the term merogony. which is used primarily to denote "multiple fission of apicomplexans" (Margulis et al. 1993). Mul- tiple fission is synonymous with progressive cleavage where re- peated karyokineses are followed by a multiple cytokinetic event yielding several daughter cells. "Successive multiple fissions" as used in Margulis et al. (1993) is considered to be a misleading term and should be avoided. The term merogony was used previ- ously for Perkinsus spp. because it is used to denote asexual cel- lular replication in the Apicomplexa to which Perkinsus spp. are related. However. 1 now believe that I did not adequately recog- nize the importance of multiple fissicni. Merogony is used to de- note one type of multiple fission in the Apicomplexa (the others being sporogony and gametogony ). Use of the term merogony was not previously considered to be a problem, because, even though cellular proliferation oi Perkinsus spp. within the ho.st most often occurs by successive bipartition. multiple fission is believed to occur infrequently (Perkins and Menzel 1967) and is known to occur in axenic cultures of P. imirinus (La Peyre et al. 1993). In addition, merogony was used previously for Perkinsus spp. be- cause the term is used in the microsporidian literature to denote cellular proliferation involving both successive bipartitioning and progressive cleavage (multiple fission) (Perkins 1991). The rea- sons I am not now using the term merogony are as follows: 1) whereas merogony is associated with both the Apicomplexa and the Microspora, Perkinsus spp. are closely related to the Apicom- plexa. not to the Microspora; therefore, the apicomplexan sense of the term (multiple fission) should dominate; 2) merozoites of the Apicomplexa (the daughter cells resulting from merogony) are motile and have an apical complex whereas the daughter cells of Perkinsus spp. are not motile and do not have apical complexes: and 3) the dominant form of cellular proliferation in Perkinsus spp. is successive bipartition not multiple fission. The cells engaged in palintomy are called tomonts which is a generalized term for a dividing cell. The term has been used in descriptions of ciliates to denote a prefission or dividing stage (Lynn and Small 1990). Despite its association with ciliates the term is being adopted due to lack of a better preexisting term. The term schizont is not used herein for tomont because the former is defined as a "multinucleate organism that will undergo schizog- ony"" (Margulis et al. 1993). and schizogony is reserved for mul- tiple fission. The cellular products of palintomy have been termed herein as trophozoites, because the term is a generalized one used for par- asitic protists. meaning the growing, feeding or trophic stage which is also an interfissional form. It is not used herein to denote necessarily the adult stage in the life cycle (Corliss and Lorn 19X5) nor is it used to denote a motile stage as defined by Margulis ct al. (1993). As part of the definition. Corliss and Lom ( 1985) did not consider it necessary for the cell to be motile. The term trophont which has the same definition is not being used here because it is more often used in place of trophozoite for nonparasitic species of protists (Corliss and Lom 1985). The term tomite has been used in descriptions of ciliates to indicate one or more products of palintomy and generally has been reserved for a stage in the life cycle which is small, tree swimmmg and nonfeeding (Margulis et al. 1993). The term could be applied to Perkinsus spp.. because the term is a general one indicating a product of division; however, since it is associated with the cili- ates. I have chosen to simply refer to the cellular products of palintomy as being immature trophozoites (the smaller cells lack- ing a large eccentric vacuole). The larger cells which are differ- entiated from the immature trophozoites and which have a large eccentric vacuole are the mature trophozoites. I'hus, immature trophozoites are the same cells which were termed merozoites and mature trophozoites are the same as meronts (Perkins 1991). The mature trophozoites of Perkinsus spp. have also been termed spores (Mackin et al. 1950). thalli or prehypnospores (Mackin 1962). mature thalli (Perkins 1969). and aplanospores (Perkins 1976b). In an earlier publication. I (Perkins 1988) used the term trophozoite as employed herein. I also continued to use the term sporangium to denote the tomont stage and the resulting comple- ment of immature trophozoites contained in a mother cell wall prior to rupture of that wall. I now recognize that the term spo- rangium is inappropriate because Perkinsus spp. are cither api- complexans or dinoflagellates. and the term is not used for either taxonomic group despite the fact that it is otherwise suitable in that it is a "hollow unicellular or multicellular structure in which propagules (cysts or spores) are produced and from which they are released" (Margulis et al. 1993). The term spore, instead of trophozoite, is not used herein be- cause it is too general in its meaning (Corliss and Lom 1985), and a more specific term is appropriate. "Zoospore"" is retained from my previous publications to de- note the bitlagellated cells or swarmers which are formed outside of the host in estuarine or sea water. This is appropriate because the term is used sometimes in the dinoflagellate literature to denote asexually produced, flagellated cells instead of the more com- monly used ""dinospores"" (Freudenthal 1962. von Stosch 1973). In addition, the term is generally used in protistological literature. Since Perkinsus spp. are closely related to the dinoflagellates, use of "zoospore" is justifled. It is not a term appropriate for the more highly evolved Apicomplexa. However, since the species of Per- kinsus comprise a phylogenctically intermediate group between the dinoflagellates and Apicomplexa. there is Justiflcation for use of the term even if those molluscan pathogens ultimately are ac- cepted as members of the Apicomplexa. the reasoning being that Structure of Phrkinshs marinus 83 it would be best to use a dinotlagellate term for the most primitive apicomplexan species. There is no comparable cellular stage m the more highly evolved Apicomplexa. The nearest cell type is the bitlugelkitcd microgamete formed by a few species of gregarines and the Coccidea (Perkms 1991 1. There is no evidence that the zoospores oi Perkinsus spp. are microgametes. The flagella of the microgametes have no mastigonemes. and there is no perforato- rium, like that found ]n microgametes. in the zoospores (Perkins 1991). The term dinospore is avoided m denoting the swarmcrs. be- cause it has not yet been demonstrated that Perkinsus spp.. al- though related to dinoflagcllates, are in fact dinoflagellates. "Swarmers"" is avoided because it is a term which is too general. The process by which zoospores are differentiated is best termed palintoniic zoosporulation (herein shortened to zoosporu- lation) instead of palintomic sporogenesis. as is used for a number of parasitic dinoflagellates (Cachon and Cachon 19871. The ad- jective palintomic is used because palintomy (successive biparti- tioning) is involved, as discussed above, for the cell stages of Perkinsus spp. found in the host. The term sporogenesis is not used, because the term spore is being avoided for reasons already noted above. In addition, in the dinonagcllatc literature, the term sporocyst is used in association with palintomic sporogenesis where the sporocysts each differentiate into dinospores (Cachon and Cachon 1987). It is best to avoid the term sporocyst since it has a very specific connotation in apicomplexan literature where it denotes a cyst containing sporozoites and formed within an oocyst. or in the gregannes it is the oocyst itself, there being no sporocyst. Oocysts develop from a zygote, and there is no evidence that zoospores of Perkinsus spp. result after karyogamy. The cell which forms zoospores is termed a zoosporangium. thus adopting another term which is used m the dinotlagellate literature to denote the cell which tonus swarmcrs (von Stosch 1973. Spero and Moree 198 1). It should be noted that 1 have avoided use of the term sporangium (see discussion on trophozoites above), because it is not used in the apicomplexan and dinotlagellate literature, but I am using the term zoosporangium for the reasons given previ- ously. Those mature trophozoites which have enlarged markedly, most notably in FTM. to form cells that are at least twice the size of the largest mature trophozoites are herein termed hypnospores. a term originally proposed by Mackin and Boswell (1956) and used by a number of investigators. In previous papers. I have used the term prezoosporangium in recognition of the fact that the cells may engage in zoosporulation. thus forming zoosporangia. I have resisted use of the term hypnospore because I ) it has been defined as a resting cyst (Taylor 1987) and has been used for a resistant stage in the life cycle where the cell can become viable after long periods of dormancy, and 2) it is a term associated with the di- noflagcllates and not the Apicomplexa. I have changed my posi- tion on the matter because Margulis et al. (1993) defined it as simply a thick-walled aplanospore with no mention of it being a resting stage. This definition, therefore, includes the hypnospores which have thick walls but which are not resistant or dormant cells in that they do not appear to survive beyond a week or two in sea water. Most die in sea water within 10 days. The judgment as to when death occurs is based on the disorganized appearance of the cells" cytological structure (Perkins, unpublished data). In addi- tion, since it is now recognized that Perkinsus spp. are closely related to the dinoflagellates. the use of a dinotlagellate term is appropriate. Furthermore, the hypnospores do not necessarily form zoospores, but may form hyphal-like outgrowths from which nonflagellated unicells emerge. Thus, the term prezoosporangium is less suitable. Morphologically defining when a mature trophozoite becomes a hypnospore is not possible because the transition is a gradual one whereby the vacuoplast disappears and the eccentric vacuole be- comes proportionately larger and is accompanied by a thickening of the cell wall. In P. mannus. the mature trophozoites average about 6 |a,m in diameter; however, zoosporulating cells have been observed which arc as small as 15 (jini. Thus, when approximately a doubling of the mature trophozoite's diameter has occurred, the transition to hypnospore is assumed to have been completed. In the case of Perkinsus sp. (probably P. allanlicus) from M. Inillhud. zoosporulation occurs in the larger tniphozoites (up to 48 |a,m in diameter; Perkins 19X8) which are isolated directly from the host tissues and v\hich have not been Ireated by FTM (Perkins 1968, Valiulis and Mackin 1969) as well as in those cells induced to enlarge in FTM (Perkins 1968). in Perkinsus sp. from M. Inillli- ica. the distinction between mature trophozoites and those cells which zoosporulate is even less well defined than in P manntis in that the eccentric vacuole in the former is often not as pronounced. The use of the term hypnospore in the former species may not be appropriate. Apparently mature trophozoites are able to ditteren- tiate directly into zoosporangia. In P. utianticus from Rudiuipes decussaius, zoosporulation was induced after treatment with FTM ( Azevedo et al. 1990). It is not known whether this occurs without such treatment. It is known that a wide diversity of Perkinsus spp. from bivalve molluscs, other than C \irf>imca. can form hypnospores as a result of treat- ment in F\M and this is followed by zoosporulation (Goggin et al. 1989. Azevedo et al. 1990, Perkins, unpublished data). Taxonomy The taxonomie position oi P. nuirinus has undergone a number of changes since first being described by Mackin et al. ( 1950) as D. marinum. They chose the name primarily because of the signet appearance of the mature trophozoites and the large eccentric vac- uole containing a vacuoplast. Mackin and Ray ( 1966) changed the name to Lahyrinllwnnxci marina as a result of observations ot "amoeboid stages in the host oyster'" and observations of pre- sumed cultures of P. marinus initiated by using hcmolyniph of infected oysters as an inoculum. They found aniocba-like Plasmo- dia with rhizoid-like mucoid processes. The plasmodia were ob- served to "segment into small spherical cells (3 to 5 |j.m) which produce mucoid tracks on which they travel in the gliding motion characteristic of spindles" (Mackin and Ray 1966). Subsequent work by myself (Perkins 1976b) and others (reviewed in Olive 1975 and Porter 1990) has shown that the Labyrinthomorpha (Page 1980) Pokorny. 1985 (syns.. Labyrinthomycota according to Por- ter 1990; Labynnthulina according to Olive 1975; Labyrinthulo- mycetes according to Moss 1991 ), of which the genus Lahynnth- omxxa was considered to be a member, are not closely related to Perkinsus spp. (Perkins 1976b). It is likely that Mackin and Ray (1966) observed cells of a labyrinthomorphid. probably Laln- rinlhulouies sp. (Perkins, 19731, which were contaminants in their cultures. The labyrinthomorphids are members of the monophyl- etic assemblage known as the siramenopiles and probably evolved early in the evolution of the group (Leipe et al. 1994). As the information accumulated concerning the characteristics of the lab- yrinthomorphids. I discontinued use of the generic name Laby- 84 Perkins rinrhimiyxa and reverted to use of the generic name Dermocystid- mm (Perkins 1976b) until I97H. Following the demonstration ot an apical complex in zoospores of the oyster pathogen (Perkins 1976a). Levine (1978) renamed the pathogen P . imihmis and established a new class Perkinsea. order Perkinsida and family Perkinsidae in the phylum Apicom- plexa in recognition of the significance of the apical complex and the uniqueness of the pathogen. He later modified the endings of the class and order names so that they became class Perkinsasida and order Perkinsorida (Levine 1988). These modifications in name endings are not accepted by some workers. Perkiii.siis is now the name used for the genus of molluscan pathogens described herein. Four species have been described, all from marine molluscs: P. inannus (Mackin, Owen and Collier. 1950) Levine. 1978 from all tissues of C. virginica: Perkinsus olseni Lester and Davis. 1981 from hemolymph. adductor muscle and mantle of the Australian blacklip abalone, Haliolis ruber: P utUmiwus Azevedo. 1989 from the gills of the Portuguese clam. R. decussaliis: and Perk- insus karlssoiu McGladdery. Cawthorn and Bradford. 1991 from tissues of the bay scallop, Argopecten irnutians. The latter species is of questionable validity, because Perkinsus spp. zoosporulation, typical of Perkinsus spp., was not observed. It was not determined whether the bitlagcllatcd cells, believed by McGladdery et al. (1991) to be zoospores, have filamentous mastigonemes and an apical complex nor was the response to FTM typical of Perkinsus spp. (McGladdery ct al. 1991 ), In the case of P. ulseni no attempts were made to determine if filamentous mastigonemes or an apical complex were present; however, typical zoosporulation was ob- served with formation of a discharge pore and tube, and the re- sponse to FTM was typical in terms of enlargement of trophozoites and staining with Lugol's iodine solution. Therefore, it is likely that the pathogen described by Lester and Davis ( 198 1 ) is a species of Perkinsus. P. allanticus has been shown to possess all charac- teristics of the genus including the formation of bitlagellatcd zoo- spores with filamentous mastigonemes and an apical complex (Azevedo el al. 1990). Whether there are numerous species of Perkuisus parasitizing marine and estuarine molluscs worldwide or only a few species is not yet clear. Using the Ray fTM technique (Ray 1952), 67 spe- cies of molluscs, mostly bivalve species, have been found to con- tain species of Perkinsus. with the distribution being from coastal temperate, subtropical and tropical waters (the possible exception being P. kiirls.mni) (Perkins 1993). It is not unreasonable to sug- gest that all species of bivalve molluscs from those coastal waters can serve as hosts of Perkinsus spp. It is uncertain how many species of Perkinsus are involved in the 67 mollusc species iden- tified thus far, particularly since Goggin et al. ( 1989) were able to demonstrate that there is a low level of host specificity for the pathogens. Using zoospores they were able to readily cross-infect from host to host. In unpublished host specificity studies con- ducted at the Virginia Institute of Marine Science using species of Chesapeake Bay molluscs, 1 have obtained results which confirm in large part the observations of Goggin et al. (1989). It is also noteworthy that the structural difterences among P. marinus, P. olseni and P allanlicus are not striking, with P. karlssoni again being the exception. The question that arises is whether the dif- ferences can be attributed to the host environment presented to the parasite. The number of species of Perkinsus parasitizing molluscs worldwide could be very small. Further transmission studies and DNA-specific probes will be concerning species identity, coupled to a fluorescent stain already provided a very usefu of Perkinsus cells, but the st; cies of Perkinsus. They have is not a member of the genus three species of Australian ni; Phytogeny useful in answering these questions In using their polyclonal antibodies , Dungan and Roberson ( 1993) have I technique for detecting the presence .lin can not distinguish between spe- provided evidence that P karlssoni nor are Perkin.sHs-\\V.Q cells found in arinc mollusc members of the genus. Recent molecular data and interpretations coupled with some of the newer morphological information described herein warrant a reevaluation of the phylogeny of Perkinsus spp. Goggin and Barker { 1993) determined the nucleotide sequence ( 1 .792 bp) for the small subunit rRNA gene of Perkinsus sp. from Anudara tra- pezia and compared it to nucleotide sequences of other organisms. From their analyses they concluded that "... Perkinsus is phy- logenetically closer to dinotlagcllatcs and to coccidean and piro- plasm apicomplexans than to fungi or flagellates." Comparisons were made to a chytrid and a species of yeast to represent the fungi and to three species of zooflagellates. Of the possibilities consid- ered, they concluded that Perkinsus sp. is most closely related to the dinoflagellate Prorocentrum mieans and the coccidean api- complexan Sarcocxslis nuiris. In another molecular study, Fong et al. ( 1993) determined the base sequence (1.793 bp) of the small subunit rRNA gene of P. marinus. Their analyses led them to conclude that "Rather than being derived from some apicom- plexan lineage, dinoflagellates are descendants of an ancestor that shares common properties with Perkinsus marinus." In a third study. Marsh et al. ( 1995) cloned and sequenced a 3,200-bp mtDNA fragment from P nuuinus that contains the 5S ribosomal RNA gene. Their phylogenetic analyses resulted in their conclusion that the oyster pathogen is more closely related to the dinoflagellates than to the Apicomplexa The morphological information contained herein leads me to conclude that Perkinsus spp. have affinities with both the Di- noflagellata and the Apicomplexa. The apical complex and micro- pores indicate affinities with the Apicomplexa. The flagellar spurs indicate affinities with the dinoflagellates in that the structures have been reported only on the transverse flagellum of three spe- cies of dintitlagellates (Dodge 1967. Leadbeater and Dodge 1967. Lee 1977) and on the shorter flagellum of Cxatliomonas truncala. a flagellate which has been classified with the cyptomonads; how- ever, the latter may not be a cryptomonad (Kugrcns et al. 1987). Mitosis in Perkinsus has not been adequately elucidated, but pre- liminary observations indicate that it is dinotlagellate-like (Dodge 1987) in that I) the nuclear envelope appears to remain intact during nuclear division; 2) deep channels, in continuity with the cytoplasm, lined bv the nuclear envelope and containing bundles of microtubules, are formed in the nucleus; and 3) kinetochore-like structures are formed on the nuclear envelope as attachment loci for the microtubules. I am developing a more complete description of the spurs and mitosis in Perkinsus spp. for publication at a later date. Obviously, more molecular and morphological information is needed from studies of the parasitic dinoflagellates and primitive Apicomplexa, such as the archigregarines, before a revision ot the classification of Perkinsus spp. is attempted. It will be important to see whether the spurs arc tound in the parasitic dinoflagellates. Structure of Perkinsus m.ajhws 85 This will require negative staining or shadow casting of whole mounts of the transverse tlagella from a number of species for transmission electron microscopic studies, not just scanning elec- tron microscopic observations, which do not yield enough resolu- tion and which have been perfonned on most of the tlagella of species of dinoflagellates characterized thus far. Parasitic di- noflagellates such as Coccidinium duhoscqui will be of particular interest in such studies. The dinospores of C. duboscqui (Chatton and Biccheler 1936) bear an interesting resemblance to zoospores of Perkinsus spp., with the transverse tlagellum of the former resembling the anterior tlagellum of the latter in the manner in which it loosely coils around the anterior third of the cell body. In addition, the dinospore cell body resembles that of Perkinsus spp. A major difference is that dinospore formation in C. duhoscqui occurs by multiple fission, not by palintomy. Another component of these considerations is the determina- tion of the taxonomic position of predatory flagellates such as Colpodella perfonins (Hollande. 19381 Patterson and Zolffel. 1991. syns. Bodo perfonins Hollande. 1938 and Spiromoiuis per- forans Brugerolle and Mignot. 1979, as well as Spiromonas gonderi Foissner and Foissner, 1984, which should be placed in the genus Colpodella. The flagellated cells of C. perforans have a three-membrane pellicle (alveolate structure), micronemes, sub- pellicular microtubules, micropores and trichocysts (Brugerolle and Mignot 1979). S. gonderi lacks trichocysts but has a conoid (Foissner and Foissner 1984). Thus, it has been suggested that the ancestor to the Apicomplexa was similar to Colpodella {Spironu>- nas) (MyPnikov 1991). Krylov and Myrnikov (1986) and Myl- "nikov (1991) noted that there are strong similarities between P. mariniis and Colpodella [Spiromonas). including a transitional cylinder (or helix?) at the base of each tlagellum and filamentous mastigonemes on the anterior tlagellum. They indicated that "it is reasonable to combine these groups." It is possible that the species oi Colpodella [Spiromonas] are predatory dinoflagellates. It would be helpful for the molecular phylogenists to evaluate these pred- atory flagellates and thus contribute to elucidation of their affini- ties. At this stage in the development of our knowledge concerning Perknisus spp.. it is interesting to note that Levine (1985) sug- gested that dinoflagellates may have given rise to the Apicom- plexa, with Perkinsus spp. possibly being the first apicomplexan to diverge from the evolutionary line which gave rise to the rest of the Apicomplexa. Future findings will probably support his sug- gestion. Based on the discovery of a 35-kb circular genome in all apicomplexans which resembles a chloroplast genome as well as the recognition that both the Apicomplexa and the Dinoflagellata are alveolates. Farmer (1995) has already stated that "the likely ancestor for the apicomplexans, a group of obligate parasites, is therefore a photosynthetic dinotlagellate." ACKNOWLEDGMENTS I am grateful to Ms. Kay B. Stubblefield and Mr. William W. Jenkins for assistance in preparation of the micrographs and draw- ings. Ms. Diane Wong-Verelle and Ms. Susan Dicus are thanked for providing expert technical assistance. 1 am indebted to Dr. Phyllis C. Bradbury for enlightening discussions and helpful sug- gestions concerning protistan temiinology, morphology and life cycles. Contribution No. 1996, School of Marine Science, Vir- ginia Institute of Marine Science, College of William and Mary. LITERATURE CITED Andrews, J. D. 1955. Notes on fungus parasites of bivalve mollusks. Proc. Nail. Shellfish. Assoc. 45:157-163. Azevedo. C. L. Corral & R Cachola. 1990. Fine structure of zoosporu- lation in Perkinsus iillanlicus (Apicomplexa: Perkinseal. Pcinisiiolo)>\ 100:351-358. Bower, S. M. 1987, Labyrimhuloides hulioiidis n. sp. (Protozoa: Laby- rinthomorpha). a pathogenic parasite of small juvenile abalone in a British Columbia mariculture facility. Can. J. Zoal. 65:1996-2007. Brugerolle. G. & J. P. Mignot. 1979. Observations sur le cycle I'ultra- structure et la position systematique do Spiromonas perforans [Bodit perforans Hollande 1938). tlagelie parasite de Chilomtnuis paramae- cium: ses relations avec les dinotlagelles et sporozoaires. Protistolog- ica 15:183-196. Cachon, J. & M. Cachon. 1987. Parasitic dinoflagellates. In: F. J. R Taylor (ed.). The Bioloi;\ of Dinoflagellates. Blackwell Scientific Pubf, Oxford, pp. 571-610. Chatton, E. 1937. Titres et Travaux Scientlt^^ue^ (1906-1937). Sottano, Sete. p. 386. Chatton. E. & B Biecheler. 1936. Documents nouveaux relatifs aux Coc- cidinides (Dinotlagelles parasites). La sexualite du Coccidinium mesmli n. sp. C. R. Acad. Sci. im.^l'i-SK). Corliss, J. O. 1994. An interim utilitanan ("user-friendly ' I hierarchical classification and characterization of the protists. Ada Prolozool. 33: 1-51. Corliss. J. O. & J. Lorn. 1985 An annotated glossary of protozoological temis. In: J. J. Lee, S. H. Hutner, & E. C. Bovee (eds). An Plus- irated Guide to the Protozoa. Society of Protozoologists. Allen Press, Lawrence. Kansas, pp. 576-602. Dodge, J. D. 1967. Fine structure of the dinotlagellate Aureodinium pig- mentosum gen. et sp. nov. Br. Phycol. Bull. 3:327-337. Dodge. J. D. 1987. Dinoflagellateultrastructure. //I. F. J. R. Taylor (ed). The Biology of Dinoflagellates. Blackwell Scientific Pubf. Oxford, pp. 92-119. Dungan. C. F. & B. S. Roberson. 1993. Binding specitlcities of niono- and polyclonal antibodies to the protozoan oyster pathogen Perkinsus marinus. Dis. Aquat. Org. 15:9-22. Farmer. M. A. 1995. Do humans want malaria? Exploring the dinotlagel- late-apicomplexan relationship. J. Phycol. 31(Suppl. to No. 3):18 (Abst.). Foissner, W. & I. Foissner. 1984. First record of an ectoparasitic flagellate on Ciliates: an ultrastruclural investigation of the morphology in the mode of attachment of Spiromonas gonderi Nov. Sp. (Zoomastigo- phora. Spiromonadidae) invading the pellicle of Ciliates of the genus Colpoda (Ciliophora, Colpodidae). Protislologica 20:635-648. Pong. D.. R. Rodriguez, K. Koo, J. Sun, M. L. Sogin, D. Bushek, D. T. J. Littlewood & S. E. Ford. 1993. Small subunit ribosomal RNA gene sequence of the oyster parasite Perkinsus marinus. Mol. Mar. Biol. Biotechnol. 2:346-350. Freudenthal. H. D. 1962. Symbiodinium gen. nov. and Symhuidinium mi- croadriaticum sp. nov. , a zooxanthella: taxonomy, life cycle, and mor- phology. J. Protozool. 9:45-52. Gauthier, J. D. & G. R. Vasta. 1993. Continuous in vitro culture of the eastern oyster parasite Perkinsus marinus. J. Invertehr. Pathol. 62; 321-323. Goggin, C. L. & S. C. Barker. 1993. Phylogenetic position of the genus Perkinsus (Protista, Apicomplexa) based on small subunit ribosomal RNA. Mol. Biochem. Parasitol. 60:65-70. 86 Perkins Goggin. C. L. & R. J. G. Lester. 1987. Occurrence of Perkinsiis species (Protozoa, Apicomplexa) in bivalves from the Great Barrier Reef. Dis. Aqiiat. Org. 3:113-117. Goggin. C. L.. K. B. Sewell & R. J. G. Lester. 1989. Cross-infection experiments with Australian Perkmsiis species. Dis. Aqua!. Org. 7: 55-59. Jones. G. M. & R. K. O'Dor. 1983. Ultrastructural observations on a thraustochytrid fungus parasitic on the gills of squid [Iltex illecebrosus Lesueur). J. Parasitol. 69:903-911. Kleinschuster. S. J. & S. L. Swink. 1993. A simple method for the in vilro culture of Perkinsiis marinus. The Nawilus 107:76-78. Kleinschuster, S. J., F. O. Perkins. M. J. Dykstra & S. L. Swink. 1994. The in vitro life cycle of a Perkinsus species {Apicomplexa. Perkin- sidae) isolated from Macoma balthica (Linneaus. 1758). J. Shellfish Res. 13:461^66. Krylov. M. V. & A. P. Myl'mkov. 1986. New taxa in the phylum Spiro- zoa. Spiromonadomorphina subcl. n.. Spiromonadida ordo n. Parazi- tologiya 20:425-430. Kugrens, P.. R. E. Lee & R. A. Andersen. 1987. Ultrastructural varia- tions in cryptomonad tlagella. J. Phycol. 23:511-518. La Peyre. J. F. & F-L. E. Chu. 1994. A simple procedure for the isolation of Perkinsus marinus merozoites. a pathogen of the Eastern oyster. Crassoslrea virginica. Bull. Eur. Assoc. Fish. Pathol. 14:101-103. La Peyre. J. F.. M. Faisal & E. M. Burreson. 1993. In vilro propagation of the protozoan Perkinsus marinus. a pathogen of the eastern oyster, Crassoslrea virginica. J. Euk. Microbiol. 40:304—310. Leadbeater. B. & J. D. Dodge. 1967. An electron microscope study of dinoflagellate flagella. J. Gen. Microbiol. 47:305-314. Lee. R. E. 1977. Saprophytic and phagocytic isolates of the colourless heterotrophic dinoflagellate Gyrodinium lebouriae Herdman. J . Mar. Biol. Assoc. U.K. 57:303-315. Leipe. D. D.. P. O. Wainright. J. H. Gunderson. D Porter. D. J. Patter- son, F. Valois, S. Himmerich & M. L. Sogin. 1994. The strameno- piles from a molecular perspective: 16S-like rRNA sequences from Labyrinthuloides niinuta and Cafeteria roenbergensis . PIncologia 33: 369-377. Lester, R. J. G. & G. H. G. Davis. 1981. A new Perkinsus species (Api- complexa, Perkinsea) from the abalone Haliotis ruber. J Invertebr. Pathol. 37:181-187. Levine, N. D. 1978. Perkinsus gen. n. and other new taxa in the protozoan phylum Apicomplexa. J . Parasitol. 64:549. Levine. N. D. 1985. Phylum II. Apicomplexa Levine. 1970. In: J. J. Lee, S. H. Hutner & E. C. Bovee (eds.). An Illustrated Guide to the Pro- tozoa. Society of Protozoologists. Allen Press. Lawrence. Kansas, pp. 322-374. Levine, N. D. 1988. The Protozoan Phylum Apicomplexa. Vol. I. CRC Press. Boca Raton. Florida, p. 9. Li, T.. B. S. Roberson & C. F. Dungan. 1994. Interannual abundance variation of environmental Perkinsus marinus cells in Chesapeake Bay waters. Natl. Shellfish. Assoc. Progr. Abst.. 86th Annual Meeting. Lynn, D. H. & E. B. Small. 1990. Phylum Ciliophora. In: L. Margulis. J. O. Corliss. M. Melkonian & D. J. Chapman (eds.). Handbook of Protoctista. James and Bartlett Publ., Boston. Massachusetts, pp. 498-523. Mackin. J. G. 1962. Oyster disease caused by Dermocystidium manniim and other microorganisms in Louisiana. Puhl. hist. Mar. Sci. Univ. Tex. 7:132-229. Mackin, J. G. & J. L. Boswell. 1956. The life cycle and relationships of Dermocystidium marinum. Proc. Natl. Shellfish. Assoc. 46:112-115. Mackin. J. G., H. M. Owen & A. Collier. 1950. Preliminary note on the occurrence of a new protistan parasite, Dermocystidium marinum n sp. in Crassostrea virginica (Gmelin). Science 111:328-329. Mackin, J. G. & S. M. Ray. 1966. The taxonomic relationships of Df r- mocystidium marinum Mackin, Owen, and Collier. J. Invertebr. Pathol. 8:544-545. Margulis, L.. J. O. Corliss, M. Melkonian & D. J. Chapman (eds.). 1990. Handbook of Protoctista. Jones and Barlett Publ.. Boston. Massachu- setts, pp. 727-807. Margulis. L., H. I. McKhann & L. Olendzenski. 1993. Illustrated Glos- sary of Protoctista. Jones and Bartlett Publ.. Boston. Massachusetts. 288 pp. Marsh. A. G.. G. R. Vasta & G. Bemardi. 1995. Cloning and sequence analysis of authentic 5S and 16S ribosomal RNA genes from a genomic library of the oyster parasite Perkinsus marinus. J. Shellfish Res. 14: 272. (Abst. from Aquaculture 1995 meeting of National Shellfisheries Association. San Diego. California. Feb. 1-4, 1995). McGladdery. S. E.. R. J. Cawthom & B. C. Bradford. 1991. Perkinsus karlssoni n. sp. (Apicomplexa) in bay xaWopi Argopecten irradians. Dis. Aquat. Org. 10:127-137. McLean. N. & D. Porter. 1987. Lesions produced by a thraustochytrid in Tritonia dioniedes (Mollusca: Gastropoda: Nudibranchia). J. Inver- tebr. Pathol. 49:223-225. Moss. S. T. 1991. Thraustochytrids and other zoosporic marine fungi. In: D. J. Patterson & J. Larsen (eds). The Biology of Free-Living Het- erotrophic Flagellates. The Systematics Assoc. Spec. Vol. No. 45. Clarendon Press. Oxford, pp. 415-426. Myl'nikov. A. P. 1991. Ultrastructure and biology of certain representa- tives of the order Spiromonadida (Protozoa). Zoologicheskiy Zhurnal. 7:5-15 (in Russian). Olive. L. S. 1975. The Mycetozoans. Academic Press. New York. pp. 21.5-241. Patterson. D. J. & M. L. Sogin. 1992. Eukaryote and protistan diversity. In: H. Hartman & K. Matsuno (eds.). The Origin and Evolution of the Cell. World Scientific. Singapore, pp. 13^6. Perkins. F. O. 1968. Fine structure of zoospores from Labyrinthomy.xa sp. parasitizing the clam Macoma balthica. Chesapeake Sci. 9:198-208. Perkins, F. O. 1969. Ultrastructure of vegetative stages in Labyrinthomyxa marina ( = Dermocystidium marinum). a commercially significant oys- ter pathogen. J. Invertebr. Pathol. 13:199-222. Perkins. F. O. 1973. A new species of marine labyrinthulid Labyrinthu- loides yorkensis gen. et spec. nov. — cytology and fine structure. Arch. Mikrobiol. 90:1-17. Perkins. F. O. 1974. Phylogenetic considerations of the problematic thraustochytriaceous-labyrinthulid-De»7«ocy.vr/(/iHm complex based on observations of fine structure. Veroeff. Inst. Meeresforsch. Bremer- haven 5(Suppl):45-63. Perkins. F. O. 1976a. Zoospores of the oyster pathogen. Dermocystidium marinum. I. Fine structure of the conoid and other sporozoan-hke organelles. J. Parasitol. bl:')S9-^lA. Perkins. F. O. 1976b. Dermocystidium marmum infection in oysters. Mar. Fish. Rev. 38:19-21. Perkins. F. O. 1976c. Fine structure of lower marine and estuarine fungi. In: E. B. Gareth Jones (ed.). Recent Advances in Aquatic Mycology. Elek Science. London, pp. 279-312. Perkins. F. O. 1987. Protistan parasites of commercially significant ma- rine bivalve molluscs — life cycles, ultrastructure. and phylogeny. Aquaculture 67:240-243. Perkins. F. O. 1988. Structure of protistan parasites found in bivalve mol- luscs. Amer. Fish. Soc. Spec. Publ. 18:93-11 1. Perkins. F. O. 1991. "Sporozoa": Apicomplexa. Microsporidia, Haplo- spondia, Paramyxea, Myxosporidia, and Actinosporidia. In: F. W. Harrison & J. O. Corliss (eds.). Microscopic Anatomy of Inverte- brates. Vol. 1. Protozoa. Wiley-Liss, New York. pp. 261-331. Perkins, F. O. 1993. Infectious diseases of molluscs. In: J. A. Couch & J. W. Foumie (eds.). Pathobiology of Marine and Estuarine Organ- isms. CRC Press, Boca Raton, Florida, pp. 255-287 Perkins, F. O. & R. W. Menzel. 1966. Morphological and cultural studies of a motile stage in the life cycle of Dermocystidium marinum. Proc. Natl. Shellfish. As.wc. 56:23-30. Perkins, F. O. & R. W. Menzel. 1967. Ultrastructure of sporulation in the oyster pathogen Dermocystidium marinum J. Invertebr. Pathol. 9: 205-229. Structure of Perkinsus marinus 87 Perkins. M. E. 1992. Rhoplry organelles of apiconiplexan parasites. Par- asitol. Today 8:28-32. Polglase, J. L. 1980. A preliminary report on the thraustochytnd(sl and labyrinthulid(s) associated with a pathological condition in the lesser octopus Etedone cinhosa. Bol. Mar. 23:699-706. Porter. D. 1990. Phylum Lahnnlhulomycota. In: L. Margulis. J. O. Cor- liss. M. Melkonian & D. J. Chapman (eds. ). Handbook ofProtoctisla. Jones and Bartlett Publ., Boston, Massachusetts, pp. 388-398. Ray. S. M. 1952. A culture technique for the diagnosis of infection with Dermocystidium mariniim Mackin. Owen and Collier in oysters. Sci- ence 116:360-361. Ray, S. M. 1954. Biological studies i;/ Dermocystidium mannum. u fun- gus parasire ofoyslers. Rice Insl. Pamph. Monogr. Biol Spec. Iss. The Rice Institute. Houston, Texas. 1 14 pp. Ray. S. M. & A. C. Chandler. 1955. Dermocystidium mannum. a parasite of oyster. Exp. Parasilol. 4:172-200. Shaw, B. L. & H. I. Battle. 1957. The gross and microscopic anatomy of the digestive tract of the oyster Crassostrea virginica (Gmelinl. Can. J. Zool. 35:325-347. Spero, H. J. & M. D. Moree. 1981. Phagotrophic feeding and its impor- tance to the life cycle of the holozoic dinoflagellate. Gymnodinium fungiforme. J. Phycol. 17:43-51. Stein, J. E. & J. G. Mackin. 1957. Some cytochemical studies of the cell wall of Dermocystidium marinum. Texas A&M Research Foundation, Project 23, Tech. Report No. 24. pp. 1-15. Taylor. F. J. R. 1987. Dinoflagellate morphology. In: F. J. R. Taylor (ed.). The Biology of Dmoflagellates. Botanical Monographs. Vol. 21 . Blackwell Scientific Publ.. Oxford, p. 71. von Stosch. H. A 1973. Observations on vegetative reproduction and sexual life cycles of two freshwater dinoflagellates. Gymnodinium pseudopalustre Schiller and Woloszynskia apiculata sp. nov. Br. Phy- col. J. 8:105-134. Valiulis. G. A. & J. G. Mackin. 1969. Formation of sporangia and zoo- spores by Lahyrinthomy.xa sp. parasitic m the clam Macoma halthica. J. Invertebr. Pathol. 14:268-270. Journal of Shellfish Research. Vol. 15. No I, 89-101. 19%, PROPAGATION AND IN VITRO STUDIES OF PERKINSUS MARINVS JEROME F. LA PEYRE School of Marine Science Viri^inia Institute of Marine Science Collciie of William and Mary Gloucester Point. Virginia 23062 ABSTRACT The development of continuous cultures of Perkinsus maniuis ( Apicomplexa) is a major breakthrough that will lead to a better understanding of this deadly oyster pathogen. More than 10 P. rnarinus isolates are currently in contmuous cultures. Culture media used to propagate P. rnarinus range from media designed for the culture of mammalian cells to protein-free chemically defined media. Continuous cultures of P. rnarinus can be initiated from a variety of infected oyster tissues or from isolated hypnospores (i.e., the enlarged stage oi P. rnarinus from oyster tissue incubated in Ray's fluid thioglycollate medium). P. rnarinus cells adapt well to culture conditions and optimal temperature, osmolality. pH and seeding density for its propagation are reported. The availability of several P. rnarinus isolates in cultures has prompted investigations that address the parasite's genetic makeup, virulence and envi- ronmental tolerance. These studies, once completed, will provide valuable insights into disease pathogenesis and host-parasite interactions. Several imponant findings have already been made in the short time since the original culture. For example, it was found that P. rnarinus secretes serine proteases that digest oyster tissues and plasma. Acid phosphatases and heat shock ("stress") proteins are also produced by P rnarinus. Moreover, it was found that P. rnarinus extracellular products suppress some oyster host defenses. Studies on the mechanisms of adaptation of P. rnarinus to environmental conditions as well as on parasite biochemistry and nutritional requirements have also begun. Finally, cultured cells are being used in screening chemotherapeutic agents for their potential use in treating infected oysters. This review is a collation of the available literature on the methods used to propagate P. rnarinus in vitro and on current investigations conducted with cultured cells. Although it is important to realize some of the limitations of in vitro studies, research using cultured P. rnarinus cells is indispensable and may lead to novel ways of controlling the parasite, KEY WORDS: Perkinsus rnarinus. oysters, protozoa, culture procedures, culture conditions, in lirro propagation INTRODUCTION The importance of developing tecliniques for the in vitro prop- agation of parasitic protozoa has long been recognized. The ability to mass culture parasites in vitro enables essential biological, met- abolic and morphological research that would otherwise be diffi- cult or impossible to accomplish. Moreover, diverse studies in- cluding in vitro drug screening, physiological, biochemical and nutritional studies can be performed without interference from the host. Hence, development of in vitro cultivation techniques is often heralded as a major breakthrough. Many techniques have been developed for the cultivation of protozoan parasites of medical and veterinary importance (re- viewed in Jensen 1983). In contrast, cultivation of protozoan par- asites that affect marine organisms, especially invertebrates, has not received as much attention. This is unfortunate since according to Sparks ( 1985) "protozoan parasites are the most common cause of disease in invertebrates," Some of these protozoans (e.g.. Per- kinsus rnarinus. Haplosporidium nelsoni. Bonamia oslrea) have been responsible for devastating mortalities in species of economic importance, including oysters. Several pioneering scientists realized the importance of prop- agating the oyster pathogen P . rnarinus in vitro and attempted to adapt the protozoan to culture conditions more than 40 years ago (Prokop 1950. Ray 1952a. Ray 1954a, Mackin 1962, Perkins 1966). There are several factors that may have contributed to the inability to establish continuous cultures off, rnarinus at the time. First, it was almost impossible to control microbial contamination since antibiotics were just becoming available by the late 1940s. Second, none of the media that were used to culture bacteria, fungi or protozoa could support the proliferation of P. rnarinus. Finally, uncertainties about the morphology of the parasite and its life stages may have contributed to improper identification of organ- isms arown in cultures from infected oysters. It was not until the early 1990s that the development of con- tinuous cultures of P. rnarinus was achieved. La Peyre et al. ( 199."?) first showed that P. nwrinus could be propagated in vitro. They discovered that continuous cultures of the parasite could be established from infected oyster heart fragments in a medium de- signed to resemble bivalve plasma composition. Following this finding and using modified commercial media. Kleinschuster and Swink (1993) isolated and propagated P. martinis from primary cultures of tissue explants of visceral ganglia, and Gauthier and Vasta (1993) initiated continuous cultures of the parasite from infected oyster hemocytes. Other researchers have since propa- gated several P. rnarinus isolates in vitro (Bushek 1994, Dungan and Hamilton 1995), This ability to propagate P. rnarinus in vitro has provided a versatile system to study the biology of the parasite. It allows, for example, investigations of P. nwrinus physiology, biochemistry, nutrition, genetics and pharmacology. Information derived through experimentation using cultured P. i?iarinus is helping to improve our understanding of the host-parasite interaction at the organismic, cellular and biochemical level. Consequently, rational methods to control this parasite may be developed. This review will summarize the methodologies used to initiate and propagate the oyster pathogen P. marmus in vitro. An overview of current investigations using in vitro culture to understand P. rnarinus dis- ease will then be presented. Also, the potential for in vitro culture to resolve issues related to the pathobiology of P. rnarinus disease will be discussed. REVIEW OF CULTURE METHODS Methodologies for the propagation of P, rnarinus have recently been described in several reports (Gauthier and Vasta 1993, Klein- schuster and Swink 1993, La Peyre et al, 1993, Dungan and Ham- ilton 1995, Gauthier and Vasta 1995, Gauthier et al. 1995, La 89 90 La Peyre Peyre and Faisal 1995b). Procedures varied greatly among re- searchers in the type of medium used, in the culture conditions, in the method used for the establishment off. marinus primary cul- tures and in the evaluation of growth rates of protozoal cells. A . Media and Initial Culture Conditions A variety of culture media have been used to propagate P. marinus (Table 1). Differences in methodology also extend to the original culture conditions such as temperature, osmolality. pH and gas phase (Table 2). 1. Serum Free Culture Media The original medium used to propagate P . marinus. designated JL-ODRP- 1 , was formulated to resemble the known composition of oyster plasma (La Peyre et al. 1993). This partly defined me- dium contains more than 80 defined constituents and is supple- mented with solutions of cod liver oil. bovine serum albumin (BSA) and yeastolate ultrafiltrate. Cod liver oil provides a rich source of lipids including long-chain io-3 polyunsaturated fatty acids such as eicosapentaenic (20;5a)3) and docosahexanoic (22: 6oj3) acids that are typically found in marine organisms (Langdon and Waldock 1981. Chu and Webb 1984). Yeastolate ultrafiltrate ( 10 kDa) provides vitamin B,t as well as nutritional peptides. BSA was initially added because of the reported beneficial properties of albumin (e.g., binding of lipids, metals and hormones with en- hanced delivery of these to the cell, detoxification and provision of additional metabolic source of amino acids) (Barnes and Sato 1980. Maurer 1992). The serum free medium JL-ODRP- 1 presents several advan- tages over commercial media in that 1 ) no animal serum is present thus decreasing considerable problems associated with the use of this complex mixture in experiments (Barnes and Sato 1980, Mau- rer 1992); 2) the salt composition and osmolality of the medium can be easily adjusted to any desired value, which is essential for studying the osmolality tolerance of P. marinus (O'Farrell et al. 1995); 3) the concentration of individual components or groups of components (e.g.. amino acids, vitamins, carbohydrates) in the media are similar to their concentrations in oyster plasma and can be independently manipulated for biochemical and physiological studies. Elimination of BSA from the culture medium JL-ODRP- 1 also enables the study of parasite-derived proteins (>10 kDa) without interference from extraneous proteins (La Peyre and Faisal 1995a, La Peyre et al. 1995a). Moreover, purification of serine proteases from conditioned medium is vastly simplified in the absence of BSA (Faisal et al. 1995a). 2. Commercial Media Supplemented with Serum Commercially available media used to propagate P. marinus include Dulbecco modified Eagle's medium (DME). DMEiHam's nutrient mixture F-12 (DME:HAM"s F-12). Leibovitz"s L-15 me- dium. NCTC-135 and RPMl-1640. supplemented with 5-20'7f fe- tal bovine serum (FBS) and/or 5-20% oyster plasma (Gauthier and Vasta 1993. Kleinschuster and Swink 1993. Dungan and Hamilton 1995). Additional ingredients, such as taurine and trehalose, which are abundant in oyster plasma, are added for growth en- hancement of P. marinus to a number of these commercial media originally designed for vertebrate cell cultures (Kleinschuster and Swink 1993. Dungan and Hamilton 1995). Proliferation of P. marinus in several of these commercial me- dia was recently compared (Dungan and Hamilton 1995. Gauthier and Vasta 1995). Dungan and Hamilton ( 1995) evaluated the pro- liferation of P. marinus in three commercial media (i.e.. NCTC- 135. RPMl-1640 and 1:1 DME:Ham's F-12) as well as in the culture medium JLP, their modification of the medium JL-ODRP- 1 . Each medium was supplemented with \0'7c FBS as well as yeast extract ultrafiltrate (0.4 mg/ml). l.-glutamine (2 niMl and a lipid mixture ( 1 % ). P. marinus growth was measured by the MTS/PMS cell proliferation assay (Cell Titer 96 AQ"'; Promega, Madison. WI). They found that proliferation of the parasite was greatest in 1:1 DME: Ham's F-12 medium. A concentration of 5% FBS in 1:1 DME:Ham"s F-12 medium was then reported to be optimal fol- lowing a comparison of various serum concentrations (Dungan and TABLE 1. Selected media and supplements used to propagate P. marinus in vitro. Medium Reference* Supplement Buffer Antibiotic JL-ODRP- 1 1:1 DME:Hams F-12 1:2 DME:Ham's F-12 L-15 (Leibovitz) NCTC-135 RPMl-1640 1:1 DME/Ham's F-12 JLP FBS \09i. oyster plasma 5% Fetuin 1.7 mg/ml FBS \OVi. oyster plasma 20%, taunne 0.5 mg/L, glucose 5 mg/L. galactose 5 mg/L. trehalose 5 mg/L, yeast extract 1 g/L. lactalbumin hydrolysate 3 g/L, (lOOx) MEM vitamin solution 10 m!/L, (lOOx) lipid mixture I ml/L FBS I0'7f , yeast extract 0.4 mg/ml, L-glutamine 2 mM, lipid mixture 1% (v/v). JL-ODRP-1 carbohydrates 1% (v/v) HEPES 25 mM.** NaHCO, 43 mM HEPES 100 mM, NaHCO, 7 mM HEPES 100 mM, NaHCO, 7 mM HEPES 25 mM, Chloramphenicol 5 ^ig/ml Penicillm G 100 U/ml, streptomycin 100 U/ml Penicillm G 100 U/ml, streptomycm 100 U/ml Penicillm G 100 U/ml, streptomycm 100 jxg/ml Penicillin G 100 U/ml, streptomycin 100 (j,g/ml * References: I , La Peyre et al. (1993); 2. Gauthier and Vasta (1993); 3, Gauthier and Vasta (1995); 4. Kleinschuster and Swink (1993); 5. Dungan and Hamilton (1995). ** 5% CO, tension. Perkinsus marinus Propagation In Vitro 91 TABLE 2. Selected culture conditions for the propagation of P. marinus in vitro. Oxmolality Temperature Basal Medium Reference* pH (mOsm/kg) (°C) Gas Phase JLODRP-l 1. 2 7.5 650 21-28°C 5<7f CO, 11 DM[£:Hanis F-12 3 7.4 -900 27°C Ambient 1;2 DMEHams F12 4 6.6 -900 28°C Ambient L-15 5 7.6 750 28°C Ambient NTCT-135 6 7.5 665 28°C Ambient RPMI-1640 6 1:1 DME/Ham's F-12 6 JLP 6 * References: 1. La Peyre et al, ( 19931, 2, La Peyre and Faisal (1995a); 3, Gauthier and Vasta (1993); 4, Gauthier and Vasta (1995); 5. Kleinschuster and Swmk (1993); 6. Dungan and Hamilton (1995). Hamilton 1995). Gauthier and Vasta (199.5) asscssecJ the effects of various FBS concentrations in conibinatmn with oyster plasma on the growth of P. marinus in the basal medium DME. 1:1 DME: Ham's F-12 or 1:2 DME;Ham's F-12. The optimal medium was found to be 1:2 DME:Ham's F-12 supplemented with 5% FBS (Gauthier and Vasta 1995). 3. Chemically Defined Media Recently, two chemically defined media have been used to propagate P. iminnus (Gauthier et al. 1995, La Peyre and Faisal submitted a). P marimis was cultured in DME:Ham"s F-12 (1:1) supplemented with 1.7 mg/ml of fetuin (Gauthier et al. 1995). Fetuin is a widely used component of culture medium (Barnes and Sato 1980). It enhances cell attachment and presumably mediates transport of nutrient!} and growth factors (Puck et al. 1967. Ab- dullah et al. 1986). The major advantage of this medium is its ease of preparation, since each ciimponent is commercially available. However, the protein fetuin could interfere with studies of para- site-derived cellular and extracellular proteins. To confirm that P. marinus growth in the fully defined medium was comparable to that in serum-supplemented media. Gauthier et al. ( 1995) followed growth over 17 days in two slightly different media-DME:Ham's 1:1 and 1:2 supplemented with either 5% FBS or 1 .7 mg/ml fetuin. They found that the stationary growth phase was reached by day 8 in all media, but that the cell populations m DVIE:Ham's 1:2 exhibited 30-40% higher optical densities than did those in DME: Hams 1:1 P. marinus has been cultured in a protein-free, chemically defined medium, in part to facilitate production of polyclonal and monoclonal antibodies against parasite extracellular and cellular proteins (La Peyre and Faisal submitted a). This protein-free de- fined medium was modified from JL-ODRP-1 medium by elimi- nation of BSA, yeastolate ultrafiltrate and cod liver oil, increasing the amino acid and vitamin concentrations and adding a defined lipid solution, as well as vitamin B,,. This new medium, desig- nated JL-ODRP-3. supported a reasonable rate of growth, with a doubling time of 18 h, and the propagation of at least six isolates of P. marinus. Moreover, two of these isolates were subcultured for at least 10 passes. Subculturing was required to permit full acclimation of the cells and to show that the defined medium supported the continuous growth of P. marinus. B. Establishment of Primary Cultures Several tissues from infected oysters were used to initiate P. marinus cultures. These included heart (La Peyre et al. 1993), visceral ganglia (Kleinschuster and Swink 1993) and hemocytes (Gauthier and Vasta 1993. Dungan and Hamilton 1995). The oys- ter heart was chosen initially by La Peyre et al. ( 1993) since this organ is relatively free of microbial contaminants compared to other oyster tissues (Hetrick et al. 1981). Excised hearts were decontaminated with a solution of several antibiotics including chloramphenicol (50 |jig/ml). gentamicin (500 |jLg/ml). kanamycin ( I mg/ml). penicillin G (1,000 U/ml). polymyxin B (500 ^.g/nil). streptomycin (1 mg/ml) and rifampicin (50 p.g/ml). Oyster- associated bacteria were found to be more sensitive to chloram- phenicol than to other antibiotics, including broad spectrum anti- biotics such as gentamicin or kanamycin. Therefore, chloram- phenicol (5 jjig/ml) was routinely added to JL-ODRP-1 medium. Gauthier and Vasta ( 1993) washed hemocytes in 4.000 U/ml each of penicillin G and streptomycin in sea water, and reported a minimum effective concentration of 100 U/ml each of penicillin and streptomycin for their medium. Likewise. Kleinschuster and Swink (1993) and Dungan and Hamilton (1995) added 100 U/ml of penicillin G and 100 |xg/ml of streptomycin to their media. It is worth noting that P . marinus has a high tolerance to four tested antibacterial agents — penicillin G (1.000 U/ml). streptomycin ( 1 .000 M-g/ml). gentamicin (5.000 jxg/ml) choramphenicol (50 |jLg/ ml) — so that they may be used at high concentrations to eliminate bacterial contaminants during primary isolation or from existing cultures (Dungan and Hamilton 1995). A simple method for initiating continuous cultures of P. mari- nus directly from the parasite, without having to establish a pri- mary culture from infected oyster tissue, was described by La Peyre (1993) and La Peyre and Faisal (1995b). This procedure consists of incubating the visceral mass of an infected oyster in Ray's fluid thioglycollate medium (RFTM) to produce large hyp- nospores (also called prezoosporangia) from the smaller histozoic stages of the protozoan. Hypnospores are then purified, decon- taminated and transferred into JL-ODRP-1 culture medium, where they divide rapidly, producing large numbers of merozoites that further proliferate. This procedure allows large numbers of cul- tured P. marinus cells to be obtained relatively rapidly from in- dividual infected oysters. C. Characterization of Cultured Celts To confirm cultured organisms were indeed P. marinus. iso- lates were characterized by ascertaining; 1 ) their morphology at the light and transmission electron microscope levels. 2) their ability to form hypnospores that stained blue-black with Lugol's 92 La Peyre solution after incubation in RFTM. 3) their reactivity to antibodies raised against P. marinus isolated from oyster tissue (i.e.. hypno- spores. merozoites) and 4) their infectivity to P. imirinus-hee eastern oysters. 1. Morphology Cells cultured in JL-ODRP- 1 with either a low concentration of BSA (<4 mg/ml) or no BSA. as well as in JL-ODRP-3 or DME: Ham's F-12 with fetuin. are morphologically similar to histozoic P. marimts (Gauthier and Vasta 1995. La Peyre and Faisal, sub- mitted a). Moreover, the diameter of the largest dividing cells seldom exceeds about 25 ixm. which is the approximate size of P. mahnus schizonts reported in vivo (Perkins 1966. 1969). The size of the smaller cultured cells (3-6 |jLm) is identical to the size of merozoites isolated from infected oysters by La Peyre and Chu (1994). Small cultured cells have prominent refractile bodies, pre- sumably lipid droplets, like those observed in isolated merozoites. In contrast, cells cultured in JL-ODRP- 1 with high (>4 mg/ml) BSA concentration, or in commercial media supplemented with FBS and/or oyster plasma, are much larger than cells found m vivo (Gauthier and Vasta 1993. Kleinschuster and Swink 1993. La Peyre et al. 1993). The presence of a high protein concentration from BSA (originally 12 mg/ml) or from FBS appears to cause enlargement of cultured cells to diameters as great as 45 |jim, which is considerably larger than cells in protein-deficient media or those observed in vivo. These larger cells acquire prominent vacuoles that occupy more than 75% of cell volume, and resemble hypnospores (Kleinschuster and Swink 1993). The ultrastructure of cultured cells is typical of P. marinas as described by Perkins (1969) (Gauthier and Vasta 1993. 1995. La Peyre et al. 1993). All of the ultrastructural details and cellular organelles observed in cultured cells, such as granular cell walls, lomosomes. vacuoles containing electron-dense volutin-like ma- terial, lipid droplets, and nuclei with prominent nucleoli contain- ing torus-shaped ribosome aggregates, are identical to those de- scribed for P. manniis. Two types of division of cultured cells have been observed: division by schizogony (Gauthier and Vasta 1993. Kleinschuster and Swink 1993, La Peyre et al. 1993) and division by binary fission or budding (Gauthier and Vasta 1993. La Peyre et al. 1993). Division by schizogony involves enlargement of cells to 20-40 [xm in diameter, depending on the type of medium, cleav- age of the cytoplasm, internal formation of daughter cells and rupture of the mother cell wall. P. marinus at various stages of schizogony have been observed in vivo (Perkins 1966. 1969). and this process is considered as the method of division of P. marinus in vivo (Perkins 1993). Divisions by binary fission or budding have also been observed in vitro by La Peyre et al. (1993) and Gauthier and Vasta ( 1993). This type of division, however, is not generally described for P. marinus (Perkins 1976. 1993) although there are some reports that it occurs in vivo (Mackin et al. 1950. Mackin and Boswell 1956). The importance of binary fission or budding in vivo is not clear. Division by budding or fission is prominent in vitro in cell populations that exhibit a high growth rate (La Peyre. personal observation). 2. Ray's Fluid Thioglycollate Test P. marinus-cuhuKd cells enlarge in RFTM and stain blue- black with Lugol's solution (Gauthier and Vasta 1993. Kleins- chuster and Swink 1993. La Peyre et al. 1993). This capacity for enlargement is characteristic of Perkinsus spp. and is the basis of a common diagnostic test for infection (Ray 1952b. reviewed by Bushek et al. 1994a). 3. Immunoassay Cultured cells are strongly positive in immunoassays with anti- Perkinsus antibody using a polyclonal rabbit anti-P. marinus se- rum raised against hypnospores. produced by Dungan and Rober- son (1993) (La Peyre et al. 1993). Likewise, Gauthier and Vasta (1993. 1995) reported cross-reactivity of protein extracts from both cultured cells and freshly isolated merozoites (trophozoites) on Western blots probed with the same antiserum, as well as with their own rabbit anti-cultured merozoite serum. 4. Infectivity Finally, the ability of cultured cells to infect eastern oysters was demonstrated by La Peyre et al. (1993). Vasta and Gauthier (1993) and Bushek etal. (1994b). La Peyre etal. ( 1993) found that P. marinus-kee Maine oysters challenged with 10'' cultured cells/ oyster by injection into the mantle cavity developed 100% prev- alences of light-intensity P. marinus infections, 8 weeks postex- posure.In contrast. Gauthier and Vasta (1993. 1995) reported heavy P. marinus infections in oysters 4 to 5 weeks following two biweekly systemic or mantle cavity injections of washed cultured cells (2 X lO'). Bushek et al. (1994b) investigated the dose re- sponse of oysters to cultured P. marinus and the fate of those cells, following exposure of oysters by three different methods (feeding, mantle cavity injection and adductor muscle injection). Only man- tle cavity or adductor muscle injections were effective for devel- opment of infections. It was suggested that most of the cultured cells were either destroyed or eliminated by the oysters since low percentages of cells were recovered 4 days postexposure. D. Measurement of Growth Rate of P. marinus In Vitro Despite all of the information on culture procedures for P. marinus in the early publications, there were few data on growth rates or optimal culture conditions. However, two recent reports on optimization of culture conditions, for the propagation of P. marinus. in commercial media supplemented with FBS. are now published (Dungan and Hamilton 1995. Gauthier and Vasta 1995). In addition, the effects of several variables such as osmolality, temperature. pH, gas phase and seeding density on the propagation of P. marinus in vitro in BSA-free JL-ODRP- 1 medium were measured (La Peyre and Faisal, submitted a). In each of these studies a different method was used to deter- mine the growth rate of parasite cells. Growth rates were measured either 1 ) by microscopically counting cells with a hemacytometer (La Peyre and Faisal, submitted a). 2) by using a tetrazolium- based cell proliferation assay (Dungan and Hamilton 1995). or 3) by measuring the optical density of cell suspensions spectropho- tometncally at 600 nm (Gauthier and Vasta 1995). Each of these methods has its advantages and disadvantages. The advantage of counting with a hemacytometer is that it provides a direct measurement of the actual number of cells. How- ever, one of the difficulties in measuring cell number is that P. marinus enlarge and can divide by schizogony causing clumps of daughter cells. Cells were disaggregated by three passages through a 25-gauge needle. Microscopic cell count was used most success- fully when the parasite exhibited high proliferation rates and di- vision was predominantly by binary fission. Using this technique, population doubling time was defined as the time required for a PERKINSUS MARINUS PROPAGATION In ViTKO 93 cell population to double its actual number, regardless of cell volume. Spectrophotometric estimation of cell proliferation by tetrazo- lium-based assay is based on reduction of sulfonated internal tet- razolium salts to colored formazan products by mitocliondnal de- hydrogenase activity in viable cells. Dungan and Hamilton ( 199.^) used a commercial cell proliferation assay (CellTiter % AQ'", Promega) which produces a soluble formazan product and thus permits direct reading of proliferation assays in microplate cultures using a microplate reader at 490 nm. This assay is rapid and convenient and inherently mcorporates viability measurements. However, this assay measures mitochondrial activity which varies as a combined function of not only cell number, but also cell size and metabolic activity. Therefore, only the relative proliferation of a specific cell population propagated under different conditions can be compared. Doubling time was defined by Dungan and Hamilton (1995) as the time required for a cell population to double its biovolume, regardless of number. Growth of P. marinus was also determined by measuring the optical density of the parasite cell suspension spectrophotomctri- cally at 600 nm (Gauthier and Vasta 1995). This technique is simple and rapid but depends not only on the number of cells but also on their sizes and degree of aggregation. All of these cell population characteristics may vary widely among cultures, ren- dering standard curve estimates of population density inaccurate. Hence, only the relative proliferation of a specific cell population propagated under different conditions can be compared. Doubling time would have to be defined as the time required for a cell population to double its biovolume. regardless of number. It is, therefore, important to keep in mind the parameter used in assessing cell growth when interpreting experimental data. It is conceivable that growth measurements obtained from either cell number, cell mitochondrial activity or optical density may differ. Ideally, a combination of a quantitative cell number technique and one that measures biovolume should be used in characterizing growth of P. marinus. Nonetheless, doubling times of P inahims have been reported (Gauthier and Vasta 1993, Dungan and Hamilton 1995. La Pcyrc and Faisal, submitted b). Gauthier and Vasta (1993) reported an estimated log phase doubling time of 24 h. Dungan and Hamilton (1995) calculated a 13-h log phase doubling time for P. marinus isolate ATCC 50439, under optimized conditions. The minimum log phase doubling time for P. marinus isolate Perkinsus- 1 was 1 7 h (La Peyre and Faisal, submitted b). Strict comparison of P. marinus growth rates is. however, not possible because of many differences between studies, such as in media, isolates, seeding densities, proliferation assays and period of cell proliferation se- lected to calculate doubling time. E. Optimization of Culture Conditions Culture conditions for P. marinus proliferation in FBS- supplemented commercial media, as well as in a protein-deficient medium (i.e.. BSA-free JL-ODRP-1). have recently been opti- mized (Dungan and Hamilton 1995, Gauthier and Vasta 1995, La Peyre and Faisal, submitted a) (Table 3). The apparent variation in the obtained results emphasizes the limitations imposed by the technique used to measure P. marinus cell growth. The culture system used for optimization varied between re- searchers. Dungan and Hamilton (1995) selected 1:1 DME:Ham's F-12 with supplements to determine the optimal temperature, os- molality and pH conditions for P. marinus. P. marinus growth was measured by the MTS/PMS cell proliferation assay. Their experiments were conducted in 96-well titer plates with a seeding density of 1 x 10" to 5 x 10" cells/ml. depending on the param- eter measured. Variable effects were usually measured 48-72 h postinoculation. The optimization experiments of Gauthier and Vasta ( 1995) were conducted in 24-well plates at 2S"C and usually at a seeding density of 1 x 10" cells/ml. They used their optimal medium. 1:2 DME:Ham"s F-12 supplemented with 5'/c FBS. Cell growth was determined spectrophotometrically by measuring the optical density at 600 nm following an experimental incubation period, generally 10 days. La Peyre and Faisal (submitted b) op- timized culture conditions in BSA-free JL-ODRP-I. Cell propa- gation was determined by measuring cell density with a hemacy- tometer. Generally, flasks (75 cm~) were seeded with 1 x 10" cells/ml and cell densities were determined on days 1, 3, 7, 9, 12 and 15 postinoculation. Temperature has the greatest effect on the growth of P. mari- nus in vitro, with an optimum at 28°C in ail of the above exper- TABLE 3. Optimization of culture conditions. Culture Condition Reference* Conditions Tested 1 4. 10, 15, 20, 25, 28, 35, 40 2 4, 20, 28, 32, 38 3 4. 17, 22, 28, 36 1 320, 480, 640, 800, 960. 1.120 3 372, 661.963, 1273 2 12, 18, 24. 30, 36. 42 1 6.0-8.5 at increment of pH 0.5 1 5.6-7.8 at increment of pH 0.2 3 7.0,7.5, 8.0 Recommended Conditions Temperature (°C) Osmolality (mOsm/kg) Salinity (ppt) pH Seeding density Optical density Cell counts (10" cells/ml) FBS concentration (%) 1,920 0.02, 0.05, 0.1, 0.15, 0.3. 0.4 0.1, 0.2. 0.4, 0.8, 1.6 0, 1, 2, 3, 4, 5, 7.5. 10. 15 0, 0.1, 1, 5, 10, 20 28 28-32 28 800 661 24-36 7.0 6.6-6.8 7.5 0.4 0.1 10 5 * References; 1, Dungan and Hamilton (1995); 2, Gauthier and Vasta (1995); 3, La Peyre and Faisal (submitted a). 94 La Peyre iments (Dungan and Hamilton 1995, Gauthier and Vasta 1995. La Peyre and Faisal, submitted b). Differing results, however, were obtained at suboptimal temperatures and were most likely due to limitations of the proliferation assays used. Dungan and Hamilton (1995) reported increasing cell proliferation with increasing tem- perature between 10 and 35°C. Gauthier and Vasta (1995) found maximum growth at 32°C; however, they reported that their cells looked less healthy than at 28°C. La Peyre and Faisal (submitted b) found that cells enlarged but did not divide at 36°C. and even- tually died. Using lower temperatures. Dungan and Hamilton (1995) reported cell proliferation at 10°C. while Gauthier and Vasta (1995) did not report significant growth between 4 and 20°C. La Peyre and Faisal (submitted a) reported that cells at 17°C were slow to enlarge but enlarged to a greater extent before di- viding, than at higher temperatures. The tetrazolium cell prolifer- ation assay was the most sensitive to changes in cell activities. These changes in cell activities, however, may or may not indicate cell propagation under suboptimal conditions (i.e.. 10 and 35°C). The optical density assay appears to be least sensitive, since no proliferation was detected for cells incubated at 20°C for 10 days. P. nuiiiiuis can be propagated in media with a wide range of osmolality (i.e. . 340-1 .920 mOsnVkg). Salinities of media or sea water solutions with osmolalities between 340 and 1 .920 mOsm/ kg. would have salinities between 12.7 and 68 ppt. From these studies it would appear that P. inurinus growth is greatest in sa- linities in the range of 24-30 ppt. La Peyre and Faisal (submitted b) found that the propagation of the parasite was greater at 661 mOsm/kg (~24 ppt) than at either 372 mOsm/kg (~ 14 ppt) or 963 mOsm/kg ( — 34 ppt). In comparison. Dungan and Hamilton (1995) reported near-optimal proliferation of P. mannas between 475 mOsm/kg (-17 ppt) and 959 mOsm/kg (-34 ppt) with a maximum at 794 mOsm/kg (—28 ppt). Moreover. Gauthier and Vasta (1995) found that growth of P. marinus was significantly lower at 18 ppt than at 24 or 30 ppt but not at 36 ppt. It is important to note, however, that the results obtained represent growth rates of cells that were acclimated to 724 mOsm/kg (La Peyre and Faisal, submitted b). 650 mOsm/kg (Dungan and Hamilton 1995) and 960 mOsm/kg (Gauthier and Vasta 1995) and then transferred to the designated osmolalities without acclimation. It is possible that the growth rate of P . mcirinus might increase at lower or higher osmolalities after a period of acclimation. Indeed we have found that cells cultured in a medium with an osmolality as low as 341 mOsm/kg (12.7 ppt) for more than a year had a growth rate similar to that of cells cultured at 724 mOsm/kg in BSA-free JL-ODRP-1 medium (La Peyre unpublished data. OTarrell 1995). P. marinus can be propagated in a wide pH range. The optimal pH for the propagation of P . marinus in vitro depended on the group of researchers but encompassed a pH range of 6.6-7.5. These pH values correspond to the reported pH range of oyster plasma (Cousserans 1975). Seeding density dramatically innuences P. marinus growth rate. In our study we found that the growth rate of cultured cells was significantly increased by decreasing the seeding density from 16 X lO*^ to 1 X lO"^ cells/ml. The optimal seeding density of lO-*" cells/ml forP. marinus is within the range (i.e.. lO^'-IO'^ cells/ml) of seeding density generally used for cell cultures (Freshney 1994). Presumably, decreasing cell density increases the amount of nutrients available per cell. In contrast. Gauthier and Vasta (1995) reported that growth rate of their cultured cells increased with inoculum size. Their original data indicate that wells receiv- ing the highest inocula had the greatest optical density after 8 days. Growth rate of the cultured cells, however, decreased with in- creasing seeding density if doubling times are calculated between initial and final optical density data. F. Additional Techniques: Mass Culture, Cloning and Cryopreservalion The scaling-up of P. marinus culture allows production of a large number of cells, as well as conditioned medium containing secreted extracellular products. P. marinus adapts well to mass culture conditions. We routinely propagate cells in large culture tlasks (75-150 cnr. 50-100 ml) and use multitray units 1 1-10 x 600 cm~. 1-10 X 200 ml; Nunc cell factories) to produce condi- tioned medium (La Peyre et al. 1995a and unpublished data). Other researchers have used roller bottles to scale up cultures (Kleinschustcr et al. 1994. Gauthier and Vasta 1995). Gauthier and Vasta ( 1995) showed that higher densities of P. marinus were accomplished in 500 ml or 1 liter of culture medium in roller bottles than in 75-cm'^ flasks containing 50 ml of culture medium. In addition to the methods mentioned, a variety of culture equip- ment for scaling up cell culture is reviewed by Griffiths ( 1992) and might be used for P. marinus mass culture. The ability to clone P. mcirinus is extremely helpful since it provides a source of genetically identical cells. Consequently, po- tential problems associated with the use of a mixed cell population are eliminated, and interpretation of experimental results from a wide range of studies (e.g.. genetic, virulence factors, physiolog- ical, biochemical) is greatly facilitated. P. marinus has been cloned by standard limiting dilution method (La Peyre et al. 1993. Gauthier and Vasta 1995). Gauthier and Vasta (1995) reported that clonal cultures were successful only when a substantial volume ( 1 : 1 ) of conditioned medium was added to fresh medium, suggest- ing that certain excreted/secreted parasite products stimulated pro- liferation. Cloning methods described for other protozoal and an- imal cells, such as cloning on semi-solid medium, might be useful for P. marinus cloning (lovannisci and Ullman 1983, Freshney 1994) but have not been reported for use with P. marinus. Genotypic alterations, as well as alterations in phenotypic ex- pression, may occur during long-term culture of any cell popula- tion. Most continuous cell strains, even after cloning, contain a range of genotypes that are constantly changing (Freshney 1994). It is thus important to be able to preserve cells early after the continuous establishment of cultures and periodically thereafter. Cryopreservation of P. marinus is done by freezing in the presence of dimethylsulfoxide (Bushek 1994, Dungan and Hamilton 1995, Gauthier and Vasta 1995). The ability of P. marinus to survive freezing was reported as early as the 1950s. Andrews and Hewatt 1 1957) froze P. marinus in oysters for several days, but failed to kill the parasite, which subsequently enlarged in RFTM upon thawing. Cultured P. marinus can also be maintained for extended periods of time (>1 year) without change of medium, or in sea water (Bushek 1994. Dungan. personal communication. La Peyre unpublished data). During this time, division of cultured cells occurs without enlargement and produces small cells of about 2 jxni. Once placed in fresh medium, these cells enlarge and prop- agate. A number of cryopreserved P. marinus isolates have been de- posited at the American Type Culture Collection (Rockville, MD) and are now available to researchers (Dungan and Hamilton 1995, Bushek personal communication). PERKINSVS MAJilMIS PROPAGATION /,V VlTRO 95 RECENT FINDINGS AND FUTURE STUDIES For the first time in over four deeadcs of P. marinus researeh we have aecess to large quantities of uncontaminuted parasite cells. Moreover, culture of the parasite provides a much simplified system to study the biology of P. marinus. independent of host influences. The success in propagating P. imiiinus in vitro has already permitted several important studies. A. Syslemalics, Genetics, Diagnostics and Detection 1 . Systematics and Genetics The combination of culture methodology and molecular biol- ogy techniques is a powerful tool to investigate inany aspects o( P. marinus biology, including its taxonomy and genetics. P. marinus ta.xonomy has always been ambiguous and is still controversial (VIvier 1982, Wolters 1991. Perkins 1996). Moreover, the dis- tinction between different P marinus races, or between P. mari- nus and other Perl^insus spp.. is not evident since there are no criteria to differentiate them morphologically (Perkins 1993). Recently, genetic molecular techniques (i.e. . polymerase chain reaction and molecular cloning) have been used to evaluate the phylogenetic position of Perkinsus spp , including P. marinus (Fong et al. 1993. Goggin and Barker 1993. Lester et al. 1993. Goggin 1994. Goggin and Cawthom 1994. Marsh et al. 1995a). The availability of uncontaminated cultured cells has been very helpful to confirm a small subunit rRNA gene sequence of P. marinus which was originally characterized from cells of infected oyster hemolymph (Fong et al. 1993). In another study. Marsh et al. (1995al. using cultured cells, reported to have cloned and sequenced a 3.2-kbp mtDNA fragment that contains the 5S ribo- somal RNA gene. Results from these studies indicate that P. mari- nus is closely related to the dinotlagellates. The existence of different races o( P. marinus could have im- portant implications for disease management but has received little attention until recently (Andrews 1955. Ray and Chandler 1955, Bushek 1992. 1994). Bushek and Allen (1994) established con- tinuous cultures of several isolates of P. mornuis from the Atlantic and Gulf coasts and compared their virulence. Isolates from the Atlantic Coast (New Jersey. Virginia) appear to be more virulent than isolates from the Gulf Coast (Texas. Louisiana). More ex- tensive studies are needed to identify races of P. marinus and to compare their virulence and environmental tolerance (Bushek 1994). 2. Diagnostics and Detection The ability to propagate P. marinus in vitro facilitates the de- velopment of DNA probes for its diagnosis in oysters and its detectionoutsideof oysters (Fong et al. 1993, Marsh et al. 1995b). Marsh et al. ( 1995b) developed a semiquantitative assay to detect P. marinus in oyster hemolymph using polymerase chain reaction amplification of an intergenic mtDNA domain of cultured P. mari- nus. Development of DNA probes may also be useful to differen- tiate between races of the parasite in oyster populations. Other specific probes such as monoclonal and polyclonal antibodies against P. marinus have been used to monitor parasite abundance in the water column (Dungan and Roberson 1993, Roberson et al. 1993). These molecular probes for P. marinus may help reveal the ecology of this parasite, as well as explain transmission dynamics. Experiments on cultured cells can also help improve existing methods that use RFI'M for P marinus diagnosis. Enlargement of cultured cells in RFTM is significant, about threefold, but is lower than expected (La Peyre et al. 1993). In vitro studies with cultured cells indicate, however, that enlargement and viability of P. mari- nus can be greatly enhanced by the addition of lipids and vitamins to RFTM (La Peyre. unpublished data). RFTM thus appears to contain necessary, but not sufficient, nutrients for the optimal enlargement of P . marinus. Hnlargement of P . marinus generally occurs when infected oyster tissue is placed in RFTM. It is thus likely that the degrading oyster tissue provides needed nutrients for the optimal enlargement of the parasite in RFTM. As a result of these studies, it was found that abundance of P. marinus cells could be increased by the addition of lipid to heavily infected oyster tissue in RFTM since the previously undetected smallest cells (l year) in what appears to be a "dor- mant" state. Division bv schizogony is considered to be the method of di- vision of P. marinus in oyster tissue (Perkins 1993). It is conceiv- able, however, that division of P. marinus by binary fission or budding may occur early in infection when the parasite is actively dividing and propagation is not nutrient limited. Division by bi- nary fission or budding may thus have been missed in histological sections since the chance of encountering the parasite in sections of oysters with light P. marinus infection is low. In addition to the production of merozoites. P. marinus can undergo zoosporogcnesis and produce zoospores. The availability of cultured cells may help elucidate some of the determining fac- tors involved in zoosporogcnesis. Zoosporogcnesis of hypnos- pores. obtained from cultured cells, has been observed with Per- kin.sus sp. originally isolated from the hard clam. Macoina Ixilth- wa. but not with P. marinus (Kleinschuster et al. 1994). It is possible that additional nutrients or specific physicochemical fac- tors are needed for inducing hypnosporc zoosporogcnesis from cultured cells. P marinus zoosporogcnesis occurs to a limited extent when hypnosporcs are isolated from infected oyster tissue incubated in RFIM and placed in culture medium (La Peyre 1993. La Peyre and Faisal 1995a). In this case, transformation of zoos- pores to merozoites has been observed. 96 La Peyre C. Physiology It is becoming clear from in vitro studies that P. mariiuis can be propagated in a wide range of environmental conditions and can survive extreme culture conditions. Among these environmental conditions, temperature and salinity are the most important in controlling the abundance of P. mariiuis in oysters. Until recendy, interpretation of results from field and laboratory studies examin- ing the effects of temperature and salinity on P. imiriiuis infection was tentative since the parasite could not be studied independent ot the host. In this context, information on the growth rate of cultured P. marinus cells at different temperatures and salinities helps ex- plain the epizootiology of the disease and may provide new ideas for management practices for disease control. Overall the results obtained in vitro are consistent with recent epizootiological and //( vivo studies. Investigations of the mechanism of adaptation of P. nuinnns to various environmental conditions may suggest novel ways to dis- rupt parasite biology in the host. Several studies have recently reported the response of P. marinus to acute osmolality and tem- perature exposure (Burreson et al. i994.0"Farrell 1995. OTarrell etal. 1995. Tirard et al. 1995). The acute osmotic tolerance of P. marinus was investigated by Burreson et al. (1994). They showed that most cultured cells are killed by acute hypo-osmotic shock although some cells are able to survive at very low osmolalities. For example 10% of cells are viable 24 h following transfer from 630 to 136 mOsnVkg. Follow- ing this initial study, cultured cells were successfully acclimated and propagated at low osmolalities (168. 341, 433 mOsm/kg; La Peyre unpublished data). Moreover, a much greater percentage (60%) of cultured cells propagated at 168 mOsm/kg survive ex- tremely low osmolality shock (56 mOsnVkg. 2.5 ppt) than cells acclimated to higher osmolalities (O'Farrell et al. 1995, OTarrell 1995). The ability of cultured cells to withstand such low osmo- lality explains recent reports on the persistence of P. marinus infection in oysters exposed to salinity lower than 3 ppt (Burreson and Ragone Calvo 1994, Ragone Calvo and Burreson 1994). The ability of P. marinus to volume regulate was recently demonstrated by OTarrell (1995). Cultured cells swelled within the first minute following a moderate hypo-osmotic shock and then returned toward original size within the next 4 minutes without lysing. Mechanisms of osmoregulation are, however, poorly un- derstood. The amino acid concentration of cells at higher osmo- lalities was not greater than that of cells at lower osmolalities (OTarrell 1995). OTarrell (1995) suggested that amino acids are not the primary osmolytes used by P. marinus in osmoregulation and that other osmolytes such as polyols may be involved. Exposure of P. marinus to acute hyperthermic shock induced the production of heat shock proteins (Tirard et al. 1995). These heat shock proteins of P. H!an;i;(.v-cultured cells differ in size and immunochemical specificity from oyster heat shock proteins. Thermal threshold for heat shock protein induction was higher in P. marinus than in oyster hemocytes. D. Virulence Factors, Infectivity and Pathogenicity 1. Virulence Factors Multiple proteases are present in P. marinus culture supema- tants (La Peyre et al. 1995a, 1995b). These proteases digest a variety of proteins including gelatin, casein, fibronectin. laniinin and plasma proteins. They belong to the serine class of proteases and are chymotrypsin-like enzymes. Protease production by P. marinus increases with increasing growth rate of the parasite at high temperature (28°C) and salinity (24 ppt) (Garreis et al. un- published data). Although the exact role(s) of P. marinus pro- teases is still unknown, several preliminary findings suggest that P. marinus proteases may play a role in invasion by the parasite ot host tissues and in counteracting both cellular and humoral de- fenses of the oyster. The finding that P. marinus proteases degrade extracellular matrix proteins (i.e.. fibronectin and laminin) may explain the extensive tissue lysis observed in heavily infected oysters and provides a possible mechanism by which P marinus can gain access to the connective tissue (Mackin 1951 . Ray et al. 1953. Ray 1954a. La Peyre et al. 1995a). The parasite body burden in oysters fed liposomes containing extracellular products (ECP) in condi- tioned medium and then challenged with P. marinus was signifi- cantly higher than that in oysters fed liposomes containing fresh medium (La Peyre et al.. submitted a) There is also an indication that P. marinus proteases sup- pressed cellular and humoral parameters of oyster host defenses (Garreis et al. 1995). Oyster hcmocyte motility was reduced by purified proteases in a dose-dependent manner. Observations ot hemocyte monolayers also indicate that P. marinus ECP enhanced degranulation of granulocytes and inhibited granulocyte motility and spreading of large hyalinocytes (La Peyre et al. unpublished data). It is possible that ECP, including proteases, may be partly responsible for the reported failure of hemocytes to successfully encapsulate groups of dividing P. marinus cells (Perkins 1976) and for the inability of hemocytes to kill and degrade or expel P. marinus at high temperature. Lysozyme activity and hemaggluti- nin titer also decrease following incubation of oyster plasma with P. marinus ECP or its purified proteases (Garreis et al. 1995). The decline of lysozyme activity and hemagglutinin titer, in vivo, have been reported in heavily infected oysters (La Peyre 1993, Chu and La Peyre 1993a). These initial studies suggest that P. marinus proteases play a significant role in the pathogenesis of the disease. Further inves- tigations on the role of proteases are needed. Moreover, the loca- tion, concentration and activity of P. marinus proteases in vivo, in infected oysters, need to be determined in order to better interpret in vitro results. Acid phosphatase is another potential virulence factor that has been measured in cell-free culture supematants (Volety and Chu 1994a. Volety 1995). Acid phosphatase activity has been located in the nucleus and the cell membrane of P. marinus as determined ultrastructurally by lead phosphate precipitation (Volety 1995). Extracellular acid phosphatase concentration in the culture me- dium is positively correlated with P. marinus cell number and temperature. Volety and Chu ( 1995) suggest that acid phosphatase may be responsible for the observed suppression of hemocyte chemiluminescence by P. marinus in response to zymosan, since other anti-oxidant enzymes, such as superoxide dismutase. cata- lase and glutathione peroxidase, were not detected in P. marinus merozoites. The significance of this finding to in vivo infections is unknown, since it has also been reported that the chemilumines- cent response to zymosan was enhanced in hemocytes collected from heavily infected oysters (Anderson et al. 1992. 1995. La Peyre 1993). The increase in hemocyte chemiluminescent re- sponse to zymosan has been attributed to hemocyte activation in diseased oysters (La Peyre 1993. Anderson et al. 1995). P. marinus produces specific proteins as a response to heat Perkinsus marinus Propagation In Vitro 97 shock (Tirard et al. 1995). Synthesis of heat shock proteins is generally induced as a response to stressful conditions. These "stress" proteins may be important for the survival of A", mm inns in oysters, especially in phagolysosomes of hcmocytes ( Lathigra et al. 1991). In this respect, the ability to differentiate heat shock proteins of P. marinus from those of oyster hemocytes as reported by Tirard et al. ( 1995) will be very useful. Further identification of P. marinus extracellular proteins and determination of their role in pathogenicity are needed. Cell sur- face protcMis also need to be mvestigated as virulence factors. For example, cell proteases and hemagglutinins are present on the surface of P. marinus and need to be characterized (La Peyre et al. unpublished data). The way in which P. marinus merozoites in- teract with the surface of oyster cells, including hemocytes. has not yet been investigated. Cell surface hydrophobicity and lectins of microbes have been implicated in a wide variety of microbial adhesion phenomena and are considered major determinants of virulence in microbial infections (Mirelman 1986. Doyle and Rosemberg 1990). 2. Infectivity and Pathugenecity The ability to mass culture P. marinus in viiro and to quantify total parasite burden provides obvious advantages for studying P. marinus pathogenesis and infection kinetics. The availability of an axenic and quantifiable inoculum for infection experiments is cer- tainly very useful for investigating the mechanisms of infection and pathogenicity of P. marinus. Caution must be exercised, how- ever, when interpreting results with cultured cells since in vitro expression of virulence factors may not entirely correspond to expression in vivo. Of major concern are possible changes in the parasite resulting from prolonged //; viiro growth. P . marinus is still infective and pathogenic after several years in culture; however, some studies suggest a loss of virulence in cultured cells (La Peyre et al. 1993, Bushek et al. 1994b. Chintala et al. 1995). In the most comprehensive study so far. the mortality rate of oysters challenged with natural parasites was found to be much greater than that of oysters challenged with cultured para- sites (75 vs 7.5%) 12 weeks postchallenge (Chintala et al. 1995). The number of cultured cells expelled (i.e., recovered from feces and pseudofeces) was nearly 20 times greater with cultured cells than with natural cells. Chintala et al. (1995) speculated that the surface of wild cells may have receptors or other characteristics that are reduced or lacking in cultured cells and that favor retention by the oyster. Results from recent infectivity studies suggest that greater prevalences and infection intensities are produced when oysters are exposed to natural P. marinus cells than similar dosage of cultured cells in analogous studies (Volety and Chu 1994b, Volety 1995). Interestingly, virulence of natural cells seems much greater in early infection studies of the 1950s (Ray 1954b. Ray and Mackin 1954. Andrews and Hewatt 1957) than it is today. The level of parasite virulence has not been duplicated in recent years except for the studies of Gauthier and Vasta (1993. 1995). The extreme virulence of Gauthier and Vasta's isolate compared to other cultured cells, as well as natural cells, if confirmed, will be useful in relating parasite characteristics (i.e.. surface and extra- cellular proteins) to virulence. Apparent differences in virulence between natural and cultured cells, as well as between cultured isolates, provide a powerful tool for analyzing virulence detenni- nants. Differences in either entry, interaction with host defenses and/or the ability to propagate in the host, between natural and cultured P. marinus cells, need to be investigated. E. Interactions With Oyster Host Defenses One requirement for a pathogen to become established in a host is that the pathogen must overcome host defenses. The pathogen may passively avoid being recognized or actively inhibit or inter- fere with host defense mechanisms. In cases where the host is able to recognize and kill pathogen cells, rapid proliferation of the pathogen in host tissues may overwhelm host defenses. This may result from: 1) infection (entry) by an overwhelming number of pathogen cells. 2) high rate of pathogen proliferation in host tis- sues or 3) depressed defense response. In the case of P. marinus. it is likely that a number of these scenarios are extant under con- ditions that favor the parasite. For example at high temperatures, propagation of cultured cells is greatly increased, secretion of factors by cultured cells with the potential to suppress cellular and humoral host defenses is increased, oyster hemocyte activities are depressed (Fisher 1988, Fisher et al. 1989, Chu and La Peyre 1993b) and natural exposure to potentially infective P. marinus is increased (Roberson et al. 1993). While P. marinus rapidly overwhelms oysters at high temper- atures (>25°C), there is increasing evidence that P. marinus can be degraded or expelled by the oyster defenses at low temperatures (<15°C). The availability of P. marinus cultures is important in this analysis since it reveals that P. marinus can still proliferate at 6 ppt and 17°C, albeit slowly, and can survive extremely low temperatures and salinities. The decrease in parasite abundance in oyster tissue in winter (Ray 1954a, Andrews and Hewatt 1957, Ragone Calvo and Burreson 1994) must therefore be due to an active elimination of the parasite by oyster host defenses. Addi- tional studies have shown that hemocytes can kill and degrade natural or cultured P. marinus cells, in vitro (La Peyre 1993, La Peyre et al. 1995c) and in vivo (Bushek et al. 1994b). respectively. Expulsion of P. marinus cells by hemocytes may also be important since hemocytes have been shown to cross epithelia and excrete undigestible biotic and abiotic materials (Stauber 1950. Tripp 1960. Alvarez et al. 1992). Further studies are needed to identify the mechanisms of killing as these may provide markers for dis- ease resistance. In addition to hemocytes. humoral factors may also play a role in oyster defenses against P. marinus infection. Unfortunately, the role of humoral factors in host defenses against P. marinus has received little attention. These humoral factors may include lysozyme, protease inhibitors and iron-binding proteins. Plasma lysozyme was proposed to be a potential defense factor against P. mrtn/!(«(Chuetal. 1993. Chu and La Peyre 1993b) since its activity is much greater in oysters maintained at low tempera- tures and low salinities, conditions that arc favorable for the elim- ination of P. marinus. In addition, preliminary studies indicate that P. marinus cells are lysed by plasma with high lysozyme activity from oysters maintained at low salinity as well as by commercial hen egg white lysozyme suspended in a low-salt so- lution (6 ppt) (La Peyre and Chu. unpublished data). Low con- centrations of sea salts were used because lysozyme activity was found to be drastically reduced at high salt concentrations. It is possible that lysozyme may accelerate P. marinus elimination at low salinity and temperature. Prevalence of infection in oysters maintained at low salinity (5 ppt) showed a more pronounced decline than the infection prevalence in oysters maintained m high La Peyre salinity (—20 ppt). during winter (Ragone Calvo and Burreson 1994). This decrease is probably due to the action of the host defenses since P. marinus in vivu or in vitro can survive such salinity (Ragone 1991, Ragone and Burreson 1993, La Peyre un- published data. OTarrell 1995). Additional preliminary studies also suggest that plasma pro- tease inhibitors may be involved in oyster resistance to P. marinus infection (Faisal et al. 1995b). Plasma protease inhibitors are present in oyster plasma from susceptible eastern oysters as well as resistant Pacific oysters (Faisal et al. 1995b). However, their ac- tivity against P. marinus proteases is greatly increased in chal- lenged Pacific oysters but not in eastern oysters (Faisal et al. 1995b). Moreover, protease inhibitors such as human a-,- macroglobulin have also been found to inhibit the proliferation of P. marinus in vitro (La Peyre et al. submitted b). Purification and identification of oyster plasma protease inhibitors active agamst P. marinus proteases may facilitate development of resistant oysters through either physiological regulation or genetic manipulation. Other plasma factors that inhibit the growth of P. marinus may include iron-binding protems. Gauthier and Vasta (1994) demon- strated that iron chelators such as transferrin inhibit the growth of cultured cells, by sequestering free iron needed by the parasite. They speculated that in vivo proliferation of P. marinus may be due in part to excess free iron, which they suggested would be available in oyster plasma during summer. Antimicrobial peptides are host defense factors that have re- cently been discovered in vertebrates and invertebrates. A number of such antimicrobial peptides (e.g., tachyplesin 1, magainins, cecropins and defensin HNPl ) have been tested against cultured P. marinus cells in vitro and can be lethal to this parasite (Morvan et al. 1995). Transfer of genes coding for antimicrobial peptides, or other resistance factors, may eventually produce resistant oysters. However, the development of techniques for transgenic technol- ogy in oysters is still limited. In conclusion, like other host-parasite systems, there is a dy- namic interaction between P. marimis and the oyster which varies with environmental conditions to favor either the host or the par- asite. It is hypothesized that the apparent ability of oysters to eliminate P. marinus at low temperatures is counteracted at high temperature by the high growth rate of the parasite and its in- creased production of virulence factors that suppress oyster de- fenses. It is evident that knowledge about the interactions of P. marinus with oyster host defenses is still very limited and further investigation is needed to test this hypothesis. F . Biochemistry, Nutrition and Chemotherapy One of the major reasons for investigating the biochemistry of P. marinus is that the results may suggest ways to control the parasite. A rational approach to chemotherapy is dependent on a working knowledge of the biochemistry of both parasite and host. Little is known, however, about the nutrition, composition and metabolism of P . marinus. The availability of large numbers of axenic parasite cells through in vitro culture has made biochemical studies on P . marinus feasible. 1. Biochemistry and Nutrition The biochemical characterization of P. marinus has just begun. The lipid and fatty acid composition of merozoites and hypnos- pores was recently characterized (Volety 1995. Volety et al. 1995). It was found that merozoites had a high percentage of phospholipids (61%) and a low percentage of triacyglycerols ( 18%) whereas hypnospores had a low percentage of phospholipid (8.3%) and a higher percentage of triacylglycerols (67%). The higher percentage of phospholipids, primary' membrane constitu- ents, may be explained by the smaller size and high proliferation rate of the merozoites compared to the much larger hypnospores which contained a high percentage of triacy glycerol, a storage lipid. Surprisingly, merozoites and hypnospores had relatively high levels of arachidonic acid (20;4n-6) compared to oysters and culture media. It is important to keep in mind that the merozoites were propagated in two different culture media supplemented with FBS. which contains animal lipids, whereas the hypnospores were isolated from infected oyster tissue following incubation in RFTM, which also contains lipids in addition to lipids from the oyster tissue. Comparison of the lipid and fatty acid compositions of merozoites isolated from oysters and merozoites propagated in defined media should provide more useful data and confirm the lipid composition of F. marinus. Interestingly, the chemically de- fined medium used by Gauthier et al. (1995) contained linoleic acid as the sole source of fatty acid for the propagation of P. marinus. Comparative analysis of input and spent defined media will be useful in determining which nutrients are utilized by P. marinus. For example, some amino acids may be selectively depleted from the medium. This simple approach could give preliminary insight into the nutritional requirements of the parasite. Moreover, it may explain some of the pathogenicity associated with P. marinus. such as its reported interference with oyster osmoregulation (Paynter et al. 1995, Paynter 1996). As would be expected, some medium components, such as soluble iron, are essential for the parasite (Gauthier and Vasta 1994). Gauthier and Vasta (1994) speculated that the increase of iron in oyster tissues in summer, possibly reflecting pollution, promotes parasite proliferation. In addition, they proposed that in hemocytes. parasites may avoid oxidative damage by depleting hcmocyte iron that is required for superoxide and hydroxyl radical production. Information gathered from nutritional and biochemical studies of P marinus is impor- tant since it may lead to novel ways of controlling this parasite. 2. Chemotherapy The availability of cultured cells has permitted screening of an array of chemotherapeutic agents for their ability to kill or inhibit the proliferation of P. marinus in vitro (Calvo 1994, Calvo and Burreson 1994, Krantz 1994, Dungan and Hamilton 1995). Prior to the development of culturing procedures, the only in vitro tech- nique available was screening the effect of drugs on P. marinus enlargement in RFTM (Ray 1966a). The relevance of enlargement of P. marinus in RFTM in relation to proliferation of P. marinus in oysters is ambiguous. Nonetheless, this technique was useful in selecting cycloheximide for (/; vivo experiments. Cycloheximide is still the only drug that has been found to reduce infection inten- sities and oyster mortality (Ray 1966b, Calvo 1994). CONCLUSION The resurgence of P. marinus as the most important oyster pathogen in the Chesapeake Bay in the mid- 1980s stimulated re- search to better understand this deadly parasite. Although numer- ous reviews discussed P. marinus morphology, life history and epizootiology prior to 1990, the lack of continuous culture of P. marinus had hindered important investigations on parasite biology and on host-parasite interaction. Fortunately, it was recently dis- covered that P. marinus could be propagated in vitro. This con- Perk/nsus marinus Propagation In Vitro 99 tinuous culture generates large quantities of uneontiwiimated P. mannus for research material and has provided a much simplitied system to study the biology of P . marinus. This new tool has opened new doors to many investigations that were cither imprac- tical, difficult or previoush impossible. Although the breakthrough in propagating P. marinus in vitro is recent, several important studies have already been accom- plished that provide important insights into disease pathogenesis, host-parasite interactions and parasite physiology. There is a need to further pursue these studies and expand on them. Particular emphasis should be placed on studying genetic and phenotypic variations between isolates and how these variations influence vir- ulence of the protozoan. It is also important to realize some of the limitations of cultured cells and to follow up in vitro findings with 1(1 vivo studies. In conclusion, research should be intensified since the devel- opment of techniques for the propagation of P. marinus is bound to yield many rewards. Investigations using cultured cells have the potential to generate new knowledge that scientists and managers may use to develop rational approaches to prophylaxis and control of this deadly oyster disease. ACKNOWLEDGMENTS The author thanks Chris F. Dungan. Dr. Mohamed Faisal, Lisa M. Ragone Calvo and Dr. Kimberley S. Reece for constructive review of the manuscript. The research was funded by a grant/ cooperation agreement from the National Oceanic and Atmo- spheric Administration, Oyster Disease Research Program, grant no. NA16FL0404-01. The views expressed herein are those of the author and do not necessarily rctlect the views of NOAA or any of Its subagencies. Virginia Institute of Marine Science contribution number 1986. LITERATURE CITED Abdullah. M.. J. A. Crowell. L. L. Tres & A. L. Kierzenbaum. mS6. Feluin: a serum component associated with rat Sertoli and spermato- genic cells in co-culture. J . Cell Physiol. 127:463^72. Alvarez, M. A.. F. E. Friedl, C. M. Hudson & R. I. ONeill 1992 Up- take and tissue distribution of abiotic particles from the alimentar\' tract of the Amencan oyster: a simulation of intracellular parasitism. J. Invcnebr. Pathol. 59:290-294. Anderson, R. S., K. T. Paynler & E. M. Burreson. 1992. Increased re- active oxygen Intermediate production hy hemocytes withdrawn from Crassostrea viri;inica infected with Perkinsii.s mariinis. Biol. Bull. 183:476-481. Anderson. R. S., E. M. Burreson & K. T. Paynter. 1995. Defense re- sponses of hemocytes withdrawn from Crassostrea virginica infected with Perkinsus marinus. J. Imertebr. Pathol. 66:82-89. Andrews. J. D. 1955. Notes on fungus parasites of bivalve molluscs in Chesapeake Bay. Proc. Natl. Shellfish Assoc. 45:157-163. Andrews. J. D. & W. G. Hewatt. 1957. Oyster mortality studies in Vir- ginia. II. The fungus disease caused by Dermocystidium mannum in oysters of the Chesapeake Bay. Ecol. Monogr. 27:1-25. Barnes. D. & G. Sato. 1980. Methods of growth of cultured cells in serum-free medium. .Anal. Biochem. 102:255-270. Burreson, E. M. & L. M. Ragone Calvo. 1994. Status of the major oyster diseases in Virginia — 1993. A summary of the annual monitoring pro- gram. Mar. Res. Rept. 93-5. Virginia Institute of Marine Science, Gloucester Point, Virginia. Burreson, E. M, L. M. Ragone Calvo, J. F. La Peyre, F. Counts & K T, Paynter. 1994. Acute osmotic tolerance of cultured cells of the oyster pathogen Perkinsus marinus (Apicomplexa: Perkinsida), Comp. Bio- chem. Physiol. l09A:575-582. Bushek, D. 1992. Perkinsus races: Management implications and initial observations of geographically distinct isolates. J. Shellfish Res. II: 190. Bushek, D. 1994. Dermo disease in American oysters: Genetics of host- parasite interactions. Ph.D. dissertation, Rutgers University. New Brunswick, New Jersey. Bushek, D. & S. K. Allen. 1994. Race-specitlc interactions between Crassostrea virginica and Perkinuss marinus. J Shellfish Res. 13:295. Bushek. D., S. E. Ford & S. K. Allen. 1994a. Evaluation of methods using Ray's fluid thioglycollate medium for diagnosis of Pcrkin.tu.s marinus infections in the eastern oyster Crassostrea virguucu Annu. Rev. Fish Dis. 4:201-207. Bushek. D., S. K. Allen, K. A. Alcox, R. Gustafson & S. E. Ford. 1994b. Dose response of the eastern oyster, Crassostrea virginica. to cultured cells of Perkinsus marinus. J. Shellfish Res. 13:313. Calvo, G. W. 1994. In vitro and in vivo effects of chemotherapeutants on the oyster parasite. Perkinsus marinus. Ph.D. dissertation. College of William and Mary. Williamsburg, Virginia. Calvo, G. W. & E. M- Burreson. 1994. In vuro and in vivo effects of eight chemotherapeutants on the oyster parasite Perkinsus marinus (Mackin. Owen and Collier). J. Shellfish Res. 13:101-107. Chintala, M. M., K. A. Alcox, S. E. Ford & D. Bushek. 1995. Cultured cells and natural Perkinsus marinus cells: A possible mechanism for virulence differences. J. Shellfish Res. 14:240. Chu. F.-L. E. & K. L. Webb. 1984. Polyunsaturated fatty acids and neu- tral liids in developing larve of the oyster, Crassostrea virginica. Lip- ids 19:815-820. Chu, F.-L. E. &J. F, LaPcyrc- i993a. Development of disease caused by the parasite. Perkinsus marinus and defense-related hemolymph factors in three populations of oysters from the Chesapeake Bay. J. Shellfish Res. 12:21-27. Chu. F.-L. E. & J. F. La Peyre. 1993b. Perkinsus marinus susceptibility and defense-related activities in eastern oysters Crassostrea virginica: Temperature effects. Dis. Aqiiat. Org. 16:223-234. Chu, F.-L. E., J. F. La Peyre & C. S Burreson. 1993. Perkinsus marinus infection and potential defense-related activities in eastern oysters, Crassostrea virginica: Salinity effects. J. Imertebr. Pathol. 62:226- 232. Cousserans, F. 1975. Recherches sur la culture de cellules de mollusques marins et sur I'emploi de ces systemes cellulaires en pathologie marine. Ph.D. these. Academic de Montpellier, Universite de Science et Tech- niques du Languedoc. France. Doyle, R. J. & M. Rosemberg. 1990. Microbial Cell Surface Hvdropho- bicity. American Society for Microbiology, Washington D.C., 425 pp. Dungan, C. F. & B. S. Roberson, 1993. Binding specificities of mono- and polyclonal antibodies to the protozoan oyster pathogen. Perkinsus marinus. Dis. Aqiiat. Org. 15:9-22. Dungan, C. F. & R. M. Hamilton. 1995. Use of tetrazolium-based cell proliferation assay to measure effects of in vitro conditions on Perk- insus marinus (Apicomplexa) proliferation. J. Eiik. Microbiol. 42: 379-388. Faisal, M., D. Y. Shafhauser & J, F. La Peyre. 1995a. Punfication of serine proteases from Perkinsus marinus using bacitracin-sepharose affinity chromatography. Aquaculture '95, February 1—4. 1995, San Diego, California. Book of Abstracts, p. 108. Faisal, M., K. Adham & J. F. La Peyre. 1995b. Inhibition oi Perkinsus marinus proteases by plasma of Crassostrea spp. Modulators of Im- mune Responses, July 8-13, Brieckenridge, Colorado, Book of Ab- stracts. Fisher, W. S. 1988. Environmental influence on bivalve hemocyte func- tion. Amer. Fish. Soc. Spec. Piibl. 18:225-237. Fisher, W. S., M. M. Chintala & M. A. Moline. 1989. Annual vanation of estuarine and oceanic oyster Crassostrea virginica (Gmelin) hemo- cyte capacity. J. Exp. Mar. Biol. Ecol. 127:105-120. Fong. D.. R. Rodnguez. K. Koo, M, L. Sogin, D Bushek, D. T. J. 100 La Peyre Littlewood & S. E. Ford. 1993. Small subunit ribosomal RNA gene sequence of the oyster parasite Perkinsus marimis. Mol. Mar. Biol. Bioiechnol. 2:346-350. Freshney. R. I. 1994. Ctillwe of Animals Cells. A Mciiuuil of Basic Tech- niques. Third edition. Wiley-Liss. Inc.. New York. 486 pp. Garreis. K. A.. J. F. La Peyre & M. Faisal. 1995. Parasitic proteases suppress immune functions in the eastern oyster (Crassosirea vir/^in- ica). Eastern Regional Symposium on Mechanisms of Immunoto.xicily. "Role of Proteases and Antigen Processing in the Immune Response." Virginia Beach. Virginia, Book of Abstract, p. 35. Gauthier. J. D. & G. R. Vasta. 1993. Contmuous in \iiro culture of the eastern oyster parasite Perkinsus marinus J . Inveriehr. Palhol. 62: 321-323. Gauthier. J. D. &. G. R. Vasta. 1994. Inhibition of in vilro replication of the oyster parasite Perkinsus marinus by the natural iron chelators transferrin, lactoferrin and dcsferrioxamine. Dev. Comp Immunol 18: 277-286. Gauthier. J. D. & G. R. Vasta. 1995 In vilro culture of the eastern oyster parasite Perkinsus marinus: Optimization of the methodology. J. In- verlebr. Palhol. 66:156-168. Gauthier. J. D.. B. Feig & G R. Vasta. 1995. Effect of fetal bovine serum glycoproteins on the in vitro proliferation of the oyster parasite Perkinsus marinus: Development of fully defined medium. J. Euk. Microhiol. 42:307-313. Goggin. C. L. 1994. Variation in the two internal transcribed spacers and 5.8S ribosomal RNA from five isolates of the marine parasites Perk- insus (Protista. Apicomplexa). Mol. Biochem. Parasitol. 65:179-182. Goggin, C. L. & S. C. Barker. 1993. Phylogenetic position of the genus Perkinsus (protista. Apicomplexa) based on small unit ribosomal RNA. Mol. Biochem. Pcirasilol. 60:65-70. Goggin, C. L. & R. J. Cawthom. 1994. Using molecules to diagnose Perkinsus species and determine the affinities of parasites of uncertain phylogenetic affinities from fish and shellfish. J. Shellfish Res. 13:296. Griffiths, B. 1992. Scaling-up of animal cell cultures. In: R. 1. Freshey (ed. ). Animal Cell Culture. A Practical Approach. IRL Press at Oxford University Press, Oxford, United Kingdom, pp. 1-14. Hetrick, F. M., E. Stephens. N. Loxmax & K. Lutrell. 1981. Attempts to develop a marine niolluscan cell line. Technical report no. UM-SG- TS-81-06. Maryland Sea Grant Program, University of Maryland. Patterson Hall. College Park. Maryland 20742. lovannisci. D. M. & B Ullman 1483. High efficiency plating method for Leishmanlu promastigotes in semidcfined or completely defined me- dium. J. Parasitol. 69:633-636. Jensen. J. B. 1983. In vitro cultivation of protozoan parasites. CRC Press. Inc.. Boca Raton, Florida. 297 pp Kleinschuster, S. J. & S. L. Swink. 1993. A simple method for the in vitro cullme of Perkinsus marinus. The Nautilus 107:76-78. Kleinschuster, S. J., F. O. Perkins. M. J. Dykstra & S. L. Swink. 1994. The in vitro life cycle of a Perkinsus species (Apicomplexa. Perkin- sidae) isolated from Macoma halihica (Linneaus. 1758). J. Shellfish Res. 13:461^64. Krantz. G. E. 1994. Chemical inhibition ot Perkinsus marinus in two ;/; vitro culture systems. J. Shellfish Res. 13:131-136. Langdon. C. J. & M. J. Waldock. I98I. The effects of algal and artificial diets on the growth and fatty acid composition of Crassostrea gi.^as spat. J. Mar. Biol. A.i.m-. U.K. 61:431-148, La Peyre. J. F. 1993. Studies on the oyster pathogen Perkinsus marinus (Apicomplexa): Interactions with host defenses of Crassostrea virs;in- ica and Crassostrea .i^i.itas and in vitro propagation. Ph.D. dissertation. The College of William and Mary. Williamsburg. Virginia. La Peyre, J. F. & F.-L. E. Chu. 1994. Simple procedure for the isolation of Perkinsus marinus merozoites. a pathogen of the eastern oyster, Crassostrea virginica. Bull. Eur. Assoc. Fish Pathol. 14:101-103. La Peyre, J. F. & M. Faisal. 1995a. Perkinsus marinus produces extra- cellular proteolytic factor(s) in vitro. Bull. Eur. Assoc. Fish Palhol. 15:28-31. La Peyre, J. F. & M. Faisal. 1995b. Improved method for the initiation of continuous cultures of the oyster pathogen Perkinsus marinus (Api- complexa). Trans. Am. Fish. Soc. 124:144-146. La Peyre. J. F.. M. Faisal & E. M. Burreson. 1993. In vitro propagation of the protozoan Perkinsus marinus. a pathogen of the eastern oyster. Crassostrea virginica. J Euk. Microbiol. 40:304-310. La Peyre. J. F.. D. Y. Schafhauser. E. H. Rizkalla & M. Faisal. 1995a. Production of serine proteases by the oyster pathogen Perkinsus mari- nus (Apicomplexa) in vitro. J. Euk. Microbiol. 42:451-458. La Peyre, J. F., D. Y. Schafhauser, K. A. Garreis & M. Faisal. 1995b. Comparison of extracellular proteins in the cell-free supematants of cultures of Perkinsus marinus isolates from the Gulf and Atlantic coasts. Aquaculture '95. February 1-4. 1995. San Diego. California. Book of Abstracts, p. 158. La Peyre, J. F., F.-L. E. Chu & W. K. Vogelbein. 1995c. In vitro inter- action of Perkinsus marinus merozoites with eastern and Pacific oyster hemocytes. Dev. Comp Immunol. 19:291-304. La Peyre. J. F. & M. Faisal. Submitted a. Development of a protein free, chemically defined, culture medium for the propagation of the oyster pathogen Perkinsus marinus (Apicomplexa). La Peyre. J. F. & M. Faisal. Submitted b. Optimal culture conditions for the propagation of Perkinsus marinus in protein deficient medium La Peyre. J, F.. H. A. Yamall & M. Faisal. Submitted a. Role of extra- cellular products in the infection of eastern oysters (Crassostrea vir- ginica) by the protozoan Perkinsus marinus (Apicomplexa). La Peyre, J. F., C. L. Sellers & M. Faisal. Submitted b. In vitro effects of a,-macroglobulin on the oyster pathogen Perkinsus marinus (Apicom- plexa). Lathigra, R B., P. D. Butcher, T. R. Garbe & D. B. Young. 1991. Heat shock proteins as virulence factors. Curr. Top. Microbiol. Immunol. 167:12.5-143. Lester. R. J. G , R D Adiard. C L. Goggin. D Hervio & E. Mialhe. 1993. DNA studies on protistan parasites of molluscs. European As- sociation of Fish Pathologists. Sixth International Conference "Dis- eases of Fish and Shellfish." p 74. Mackin. J. G. I95I. Histopathology of infection of Crassostrea virginica (Gmelin) by Dennocystidium marinum Mackin, Owen and Collier. Bull. Mar. Sci. GulfCarrih. 1:72-87. Mackin. J. G. 1962. Oyster diseases caused by Dermocvslulium manniim and other organisms in Louisiana. Piibl. Inst. Mar. Sci. Univ. Tex. 7:132-229. Mackin. J. G. & J. L. Boswell. 1956. The life cycle and relationship of Dermocysiulium. Proc. Natl. Shellfish Assoc. 46:112-115. Mackin. J G . H M. Owen & A. Collier. 1950. Preliminary note on the occurrence of a new protistan parasite. Dermocystidium marinum n. sp. in Crassostrea virginica (Gmelin). Science 1 1 1:328-329. Marsh. A. G.. G. R. Vasta & G. Bemardi. 1995a. Cloning and sequence analysis of authentic 5S and 16S ribosomal RNA genes from a genomic library of the oyster parasite Perkinsus marinus. J . Shellfish Res. 14: 272. Marsh. A. G.. J. D. Gauthier & G. R. Vasta. 1995b. A semiquantitative PCR assay for assessing Perkinsus marinus infections in the eastern oyster. Crassostrea virginica. J. Parasitol. 81:577-583. Maurer. H. R. 1992. Towards serum-free chemically defined medium for mammalian cell culture. In: R. I. Freshney (ed.). Animal Cell Culture. A Practical Approach. IRL Press at Oxford University Press. Oxford. United Kingdom, pp 15^6. Mirelman. D. 1986. Microbial Lectins and Agglutinins. Properties ami Biological Activitx. John Wiley and Sons Inc.. New York. 443 pp. Morvan. A.. J. -P. Cadorel, S. Iwanaga & E. Bachere. 1995. Increasing disease resistance in marine bivalves using antimicrobial peptides: //) vitro investigations on peptide efficiency and selectivity against oyster pathogens. European Association of Fish Pathologists Seventh Inter- national Conference "Diseases of Fish and Shellfish." Palma de Mal- lorca. Spain. Book of Abstracts, p. 18. O'Farrell. C. L. 1995. Osmotic tolerance and volume regulation in cul- tured cells of the oyster pathogen Perkinsus marinus. M.S. thesis. College of William and Mary. Williamsburg, Virginia. Perkinsus marinus Propagation In Vitro 101 O'Fan-ell. C. L., J. F. La Peyre & E. M. Burreson. 1995. Acute osmolic tolerance of cultured cells of the oyster pathogen Perkinsus marinus acclimated to low salinity. J. Shellfish Res. 14:274. Paynter. K. T. 1996. The effects of Perkinsus marinus infection on phys- iological processes of the eastern oyster, Crassostrea virginica. J. Shellfish Res. 15:119-125. Paynter. K. T.. S. K. Pierce & E. M. Burreson. 1995. Levels of intra- cellular free amino acids used for salinity tolerance by oysters iCras- soslrea virginica) are altered by protozoan iPerkinsus marinus) para- sitism. Mar. Bull. 122:67-72. Perkins, F. O. 1966. Life history studies of Dermocysiulium marinum. an oyster pathogen. Ph.D. dissertation, Florida State University, Talla- hassee, Florida. Perkins, F. O. 1969. Ultrastructurc of vegetative stages in /.uMnHf/iomv.w marina ( =Dermocystiilium marinum). a commercially significant oys- ter pathogen. J. Inverlehr. Pathol. 13:199-222. Perkins, F. O. 1976. DermocMlidium marinum infection in oysters. Mar. Fish. Rev. 38:19-21. Perkins, F. O. 1993. Infectious diseases of mollusks. In: J. A. Couch & J. W. Foumie (eds.l. Pathohiology of Murine and Esiuarine Organ- isms. CRC Press, Boca Raton. Florida, pp. 255-2S7, Perkins, F. O. 1996. The structure o( Perkinsus marinus (Mackin, Owen & Collier. 1950) Levine, 1978 with comments on taxonomy and phy- logeny of Perkinsus spp. J. Shellfish Res. 15:67-87. Prokop, J. F. 1950. Infection and culture procedures employed in the study of DermocMiilium marinum. Report submitted to Texas A&M Research Foundation, Galveston. Texas (mimeographed). Puck, T. T., C. A. Waldrcn & C. Jones. 1967. Mammalian cell growth proteins, 1. growth stimulation by fetuin. Proc. Nail. Acad. Sci. USA 59:192-199. Ragone, L. M. 1991. The effect of low salinity on established infections of Perkinsus marinus in the eastern oyster, Crassostrea virginica. M.S. thesis. College of William and Mary, Williamsburg, Virginia. Ragone, L. M. & E. M. Burreson. 1993. Effect of salinity on infection progression and pathogenicity of Perkinsus marinus in the eastern oys- ter, Crassostrea virginica (Gnielinl. J. Shellfish Res. 12:1-7. Ragone Calvo, L. M. & E. M. Burreson. 1994. Characterization of over- wintenng infections of Perkinsus marinus (Apicomplexa) in Chesa- peake Bay oysters. J Shellfish Res 13:123-130. Ray, S. M. 1952a. Cultural studies of Dermocystidium marinum with special reference to diagnosis of this parasite in oysters. M.S. thesis. The Rice Institute, Houston, Texas. Ray, S. M. 1952b. A culture technique for the diagnosis of infections with Dermocystidium marinum. Mackin, Owen, and Collier, in oysters. Science 116.360 Ray, S- M. 1954a. Biological studies of Dermocvstidium marinum. Rice Institute Pamphlet, Special Issue. The Rice Institute. Houston, Texas, November, 1954, 114 pp. Ray, S. M. 1954b. Experimental studies on the transmission and patho- genicity of Dermocystidium marinum. a fungous parasite of oysters. J. Parasitol. 40:235. Ray, S M. 1966a Cycloheximide: Inhibition of Dermocystidium mari- num in laboratory stocks of oysters. Proc. Natl. Shellfish. Assoc. 56: 31-36. Ray, S. M. 1966b. Effects of various antibiotics on the fungus Dermo- cystidium marinum in thioglycollale cultures of oyster tissues. J. In- vertehr. Pathol. 8:43,^^38. Ray, S. M. & J. G. Mackin. 1954. Studies on the transmission and patho- genicity of Dermocystidium marinum. A report from the Texas A&M Research Foundation, College Station, Texas, Technical Report No. 11, Project No. 23. Ray, S. M. & A. C. Chandler. 1955. Dermocystidium marinum. a parasite of oysters. Exp. Parasitol. 4:172-200. Ray, S M., J. Mackin & J. L. Boswell. 1953. Quantitative measurement of the effect on oysters of disease caused by Dermocystidium marinum. Bull. Mar. Sci. GulfCaribb. 3:6-33. Roberson, B. S., T. Li & C. F. Dungan. 1993. Flow cytometric enumer- ation and isolation of immunonuorescent Perkinsus marinus cells from estuarine waters. J. Shellfish Res. 12:138. Sparks, A. K. 1985. Synopsis of Invertebrate Pathology E.xclusive of In- sects. Elsevier Science Publishers B.V,, Amsterdam. Stauber. L. A. 1950. The fate of India ink injected intracardially into the oyster. Oslrea virginica (Gmelin). Biol. Bull. 98:227-241. Tirard. C. T.. R. M. Grossfeld. A K. Volety & F.-L. E. Chu. 1995. Heat shock proteins of the oyster parasite Perkinsus marinus. Dis. Aqiiat. Org. 22:147-151. Tripp. M. R. 1960. Mechanisms of removal of injected microorganisms from the American oyster. Crassostrea virginica (Gmelin). Biol. Bull. 119:210-223. Vivier. E. 1982. Reflexions et suggestions a propos de la systematique des sporozoaires: Creation d'une classe des Hematozoa. Protistologica 18: 449^57. Volety. A. K. 1995. A study of the histozoic oyster parasite. Perkinsus marinus: I) Disease processes in American oysters {Crassostrea vir- ginica); II) Biochemistry of Perkinsus marinus. Ph.D. dissertation. The College of William and Mary. Williamsburg. Virginia. Volety. A. K. & F.-L. E. Chu. 1994a. A comparative study of acid phos- phatase in the parasite Perkinsus marinus and its host Crassostrea virginica. J. Shellfish Res. 13:297. Volety. A. K. & F.-L. E. Chu. 1994b. Conipanson of infectivity and pathogenicity of meront (trophozoites) and prezoosporangia stages of the oyster pathogen Perkinsus marinus in eastern oysters. Crassostrea virg//i(ca (Gmelin, n9\). J. Shellfish Res. 13:521-527. Volety, A. K. & F.-L. E. Chu. 1995. Suppression of chemiluminescence of eastern oyster {Crassostrea virginica) hemocytes by the protozoan parasite Perkinsus marinus. Dev. Comp. Immunol. 19:135-142. Volety, A. K., F.-L. E. Chu & S. Ozkizilcik. 1995. Biochemical charac- terization of the oyster parasite, Perkinsus marinus: Lipid and fatty acid composition. J. Shellfish Res. 14:280. Wolters, J. 1991 . The troublesome parasites: Molecular and morphological evidence that apicomplexa belong to the dinoflagellate-ciliate clade, Bio.fystems 25:75-83. Joiiiiml of Shellfish Rfsciinb. Vol, 15. No. I. 103-107. 1996. RACES OF PERKINSUS MARINUS DAVID BUSHEK' AND STANDISH K. ALLEN, JR.^ ^Banich Marine Field Laboratory Universin- of South Carolina Georgetown, South Carolina 29442 ^Haskin Shellfish Research Laboratory Rutgers University Port Norris. New Jersey 08349 ABSTRACT The existence of parasite races is an integral component of host-parasite interactions with significant implications for host-parasite coevolution. ecology, and management. Despite nearly 50 years of research, few studies have considered the existence or implications of races of Perkinsti.t imirinus. Nonetheless, several field and laboratory observations indicate races exist that vary in virulence or environmental tolerance. One of the keys to understanding and managing P mciriniis lies in the identification and characterization of races. A related and equally important key is elucidation of its genetic population structure. This paper discusses our current knowledge concerning P. mariniis races and population structures. KEY WORDS: Demio. disease, host-parasite interaction, resistance, virulence. Crassostrea virginica. population genetics INTRODUCTION Perkinsus marinus has long been recognized as a serious oyster pathogen that is often blamed for widespread mortaUty of the eastern oyster Crassostrea virginica (see Ray 1996. Andrews 1996). The economic and ecological value of the eastern oyster has motivated numerous studies on this parasite, but few have considered the existence of races. Races are phenotypically and genetically distinct groups of conspecific individuals (King and Stansficld 1990). Their existence is an important component in the genetics and evolution of host-parasite interactions. Identification and characterization of pathogen races have played an important role in protecting agricultural crops from disease (Day 1974) and might also be important in protecting populations of the eastern oyster from P. marinus. Any trait can be used to distinguish races, but we arc primarily interested in environmental tolerance and virulence because these traits have economic and ecological significance — they determine the impact of a particular race on host populations. Mackin ( 1962) conceded that there may be races of P. nuiriniis. but in the absence of clear evidence, he assumed that all variation in virulence was a result of environmental variation. This same assumption on the part of other researchers may be one reason that few studies have considered the implications of P. marinus races. Another reason for the lack of studies on pathogen races may be technological, specifically, the logistics of acquiring and sustaining races of the parasites for disease transmission. Recently, in vitro culture meth- ods for P . marinus have been developed (La Peyre et al. 1993, Kleinschuster and Swink 1993. Gauthier and Vasta 1993). In this paper, we review information on the population structure of P . marinus and evidence for the existence of P. marinus races and discuss the implications of races for managing oyster populations. POPULATION STRUCTURE AND THE FORMATION OF RACES The term "population" is often used in many different con- texts. It always refers to a group with something in common. To an ecologist, a population is simply a group of conspecific indi- viduals that share a common geography, it is defined as those individual living in a particular space at a particular time (Krebs 1985). To a geneticist, a population is a group of conspecific individuals that share a common gene pool (King and Stansfield 1990). Thus, genetic populations are defined by the extent of gene flow (i.e., the exchange of genetic information) between groups of individuals. Restrictions in gene flow between populations, i.e., genetic isolation between populations, may lead to the formation of races via random genetic drift or differential selection. Gene flow normally occurs when individuals migrate from one popula- tion to another. Hence, population structure plays an important role in racial development. Several levels of population structure can be defined for P. marinus based upon its ecological distribution. Originally discov- ered off the coast of Louisiana (Mackin et al. 1930)/'. marinus has been found from Massachusetts to Florida along the Atlantic Coast and from Florida to Texas along the Gulf Coast (Andrews 1988, Ford 1992). Its extension south of the United States is poorly documented, but Burreson et al. ( 1994) have described a P. mari- nus-l\kc parasite from oysters along the Yucatan coast of Mexico. Infected oysters can harbor more than 10 million P. marinus g~ ' wet tissue weight and virtually 100% of the adult oysters on a bed may be infected year-round (Bushek et al. 1994). Infective stages are transmitted between oysters through the water column (Ray 1954), where concentrations as high as 1 .9 x 10"* cells L ' have been reported (Dungan and Roberson 1993). Other vectors of transmission have been described (White et al. 1987), but direct water-borne transmission is probably the predominant mechanism. Because they are filter feeders, oysters subsample the planktonic infective stages during feeding and respiration. Hence, each oyster is probably infected with multiple clones of the parasite — sexual reproduction of P. marinus has not been observed (Levine 1988) — and the parasites within an infected oyster probably rep- resent clones from a single population. Parasites in adjacent oys- ters, on different oyster beds, in separate tidal creeks, in distinct estuaries, or on different coasts (i.e., Atlantic versus Gulf) repre- sent various levels of P. marinus population structure. From a genetic standpoint, there may be no difference among populations at any of these levels. Gene flow within, and possibly among, populations of P. mari- nus occurs during transmission of the parasite. Natural dispersal distances during transmission are unknown. Epizootiological data and transmission experiments indicate that distances as short as 15 103 104 BUSHEK AND ALLEN m can significantly reduce rates of transmission (Andrews and Hewatt 1957). This may approximate a dispersal distance limit, but the reduction is more likely due to dilution of the propagules as they move away from their source {Mackin 1962). Nonetheless, the probability that an infective cell will be transmitted from one oyster to another decreases as distance between the oysters in- creases. Distance represents a restriction to gene flow as fewer and fewer clones are transmitted between oysters. It follows that the probability that two oysters are infected with the same array of P. morinus clones (i.e.. the same population) also decreases as dis- tance increases. The physical separation of estuaries represents a second potential restriction in gene flow. We call these "potential restrictions" because at the present time we have no idea of their effectiveness. Considering both possibilities and the extensive range off. marinus. genetic isolation is likely among many pop- ulations. The stronger the isolation, the greater the chance races will form. Populations lacking gene flow may differentiate due to genetic drift or natural selection. It is unlikely that genetic drift will lead to races due to the enormous population sizes of P. marinus. A more probable cause is differential selection, which may result from environmental variation, variation in host resistance, or in- traspecific competition. For environmental causes, temperature and salinity appear to be the two most important factors governing the distribution and abundance of P. marinus. Below about 15°C, epizootics go into remission (Andrews and Hewatt 1957). In the field, prevalence tends to decline with decreasing salinity (Mackin 1956, 1962, Soniat 1985. Soniat and Gauthier 1989) and salinity below 12 ppt retards the development of infections in the labora- tory (Ragone and Burrcson 1993). Both factors vary latitudinally and have resulted in physiological races of C. virginica (Stauber 1950) that appear to be genetically distinct (Barber et al. 1991). Variation in host resistance may cause variation in the parasite by selecting various phenotypes that complement host resistance mechanisms. Finally, clones may compete for susceptible hosts or nutrients within hosts. Those with the highest rates of transmission and/or proliferation should be competitively superior. Forces also exist that counteract the mechanisms of isolation described above. Historically, oystermen and fishery management programs have transplanted oysters within and between estuaries. These actions have probably moved infected oysters, mixing what may have once been isolated populations of P. marinus. Despite such movements, some semblance of the overall population struc- ture has probably remained due to the extensive range of the par- asite and the limited movements of oysters. Oysters have not been spread up and down the coasts haphazardly. Rather, most trans- plantation has occurred within and between nearby estuaries in a somewhat controlled effort to restock dwindling populations, or to protect snrviving oysters by moving them into areas less favorable for P. marinus (i.e.. lower salinity tributaries). Most states now restrict or prohibit importation of oysters. Movement of oysters within or between adjacent estuaries may have destroyed any local population genetic structure, but there is a good possibility that population structure at a larger, regional scale (e.g.. Gulf versus Atlantic or mid-Atlantic versus south- Atlantic) has remained in- tact. The population genetic structure of P. marinus is currently under investigation (NOAA 1995). EVIDENCE FOR RACES OF P. MARINUS Several field observations indicate that races may vary in en- vironmental tolerance, virulence, or both. Despite the positive correlations between salinity and infection prevalence and inten- sity (Mackin 1956. 1962. Soniat 1985. Soniat and Gauthier 1989). high levels of infection persist in some low-salinity bays along the Gulf Coast of the United States (Craig et al. 1989. Wilson et al. 1990). In Chesapeake Bay. P. marinus spread from high- to low- salinity areas during the 1980s (Burreson and Andrews 1988). The persistence of P. marinus in low-salinity bays in the Gulf and its spread up Chesapeake Bay may indicate the existence of races tolerant to low salinity. Burreson and Andrews (1988) also re- ported that the rate P. marinus spread from one bed to another increased in Chesapeake Bay during the 1980s. This may denote the occasion of a more virulent race. Finally, the range of P. marinus has expanded northward. Prior to 1990, accounts of P. marinus in Delaware Bay were rare (Ford 1992, 1996), with out- breaks of disease usually associated with the importation of south- em oysters. Generally, the parasite disappeared after winter, pre- sumably because it could not tolerate the colder winters of the more northern latitude. The recent spread of P. marinus into Del- aware Bay. and further north along the Atlantic Coast, may rep- resent a cold-tolerant race. All of this information is. however, circumstantial. As an alternative explanation for the northward spread of P. marinus (and perhaps the other range expansions), Ford (1992) hypothesized that climatic changes, specifically a warming trend, may have created a more hospitable environment for the parasite. Differentiating between these hypotheses will re- quire identification of the genetic population structure of P. mari- nus and characterization of isolates from distinct locales with re- spect to virulence and environmental tolerance. To date, only two studies have compared geographically dis- tinct isolates of P. marinus. The primary reason for the lack of work in this field has been the inability to obtain pure isolates of the parasite that can be manipulated under laboratory conditions. Recent development of in vitro culture methods for P. marinus (La Peyre et al. 1993. Kleinschuster and Swink 1993. Gauthier and Vasta 1993) has overcome this problem. Before P. marinus could be cultured. Perkins and Menzel (1966) compared isolates of P. marinus directly. They found no difference in the ability of biflagellated zoospores from two iso- lates to infect excised tissue. Because of their motility and pos- session of an apical complex, billagellated zoospores are possibly the primary infective stage during water-borne transmission, but naturally occurring zoospores remain undescribed. They are formed after infected tissue has been incubated in Ray's fluid thioglycollate media (RFTM) and the enlarged parasites trans- ferred to sterile seawater (Perkins and Menzel 1966). Interestingly. Perkins and Menzel ( 1966) reported detecting few foci of infection in excised tissues exposed to bitlagellated zoospores even though explants probably lack many of the defenses which a parasite must overcome. Apparently, the zoospores were uninfective under the conditions of the experiment. The role of zoospores remains un- clear and recent attempts to produce P. marinus zoospores using identical or similar techniques have failed (Bushek 1994 and per- sonal communications with F-L. Chu, C. F. Dungan, S. J. Klein- schuster, and F. O. Perkins). Similar techniques have worked well to produce infective zoospores in other Perl z :c :£ CO 2i C w 2 z - (A(lanlic) South Fast — ((.uin "- West) (.KX.RAPMK ISOLATE Figure 1 . Effect of parasite origin on enlargement of P. marinus iso- lates in RFTM (Ray 1966) at 20°C. The x-axis indicates isolate of the parasite southward along the Atlantic Coast and westward along the Gulf Coast. Dashed line represents apparent clinal break. Isolate ab- breviations: DBNJ = Delaware Bay NJ: CRMD = Choptank River MD; MBVA = Mobjack Bay VA; NRNC = Neuse River NC: CPSC = Cherry Point SC: SIGA = Skidaway Island GA: SRFL = Sebastian River FL; TBFL = Tampa Bay FL; BBLA = Barataria Bay LA: SLTX = Sabine Lake TX; GBTX = Galveston Bay TX; LMTX = Laguna Madre TX. Reproduced from Bushek ( 1994). 2). These results provide a clear demonstration of differences in virulence among isolates of P. marinus. indicating the likely ex- istence of races. Furthermore, it may be possible to relate specific biochemical properties of these isolates to their virulence. Faisal et al. (1994) detected extracellular proteases, which may represent virulence factors, in the supemates of P. marinus cultures. It would be interesting to relate protease production among isolates of P. marinus to their infectivity. Undoubtedly, many future stud- ies will use tn vitro cultured parasites to compare and characterize distinct isolates of P. marinus. INTERACTIONS AMONG HOST-PARASITE RACES We have considered the formation of P . marinus races, but races of C. virginua that vary in their response to P. marinus must also be considered (Bushek 1994). Parasite virulence and host resistance are dependent upon each other, i.e., the genetic basis of one cannot be determined without considering the other (Nelson 1973, Day 1974). For example, apparent differences in virulence of P. marinus from different regions may actually reflect differ- ences in host resistance. To be certain that differences are due to parasite virulence, hosts of the same or similar genetic makeup must be used and exposed to parasites under identical conditions. The converse may also be true; apparent differences in host resis- tance may actually reflect differences in parasite virulence. The interaction can be further complicated by environmental effects. Only one study has examined racial interactions between P. marinus and C. virginica. By separately exposing offspring from four oyster populations to four isolates of P. marinus, Bushek and Allen (1996) demonstrated variation in resistance among oyster populations and variation in virulence among parasite isolates. The experimental design also enabled them to detect any genetic in- teraction. Intuitively, one may expect a "race-specific" interac- tion where each host race is resistant to a specific parasite race and each parasite race is virulent to a specific host race. For example. Galveston Bay oysters may be more resistant to Galveston Bay P. marinus than Chesapeake Bay P. marinus because they have had more time to respond to Galveston Bay P. marinus. However. Bushek and Allen (1996) found no interaction. Resistance and virulence were "general." not "race-specific." In other words. 35 30 ^ ^H ^H 25 - I 1 20 - ■ ■ 15 - ■ ■ to ■ 5 - i 0 s * S M i ■«; E c il DBNJ MBVA BBLA LMTX Atlantic Gulf Figure 2. Geometric mean infection intensity of oysters 94 days posti- noculation with four different isolates of in vi/ro-cultured P. marinus. A planned comparison between Atlantic and Gulf isolates indicated that the Atlantic isolates produced significantly heavier infections (p = 0.013). Isolate abbreviations as in legend to Figure 1. Reproduced from Bushek (1994). 106 BUSHEK AND AlLEN the most virulent parasite isolate in one oyster population remained the most virulent across all oyster populations. Similarly, the most resistant oyster population to any isolate of P. marimis remained the most resistant to all parasite isolates. This experiment was conducted under one set of environmental conditions (27°C and 25 ppt salinity) and there is a possibility that changes in these envi- ronmental conditions may lead to a different outcome. Particu- larly, we note that different environmental conditions include the variation in the isolation and culture of the parasite itself. That is. these results may have been specific to the changes in virulence that may have occasioned the in vitro culture of P. mariiuts. Fur- ther investigations are needed to determine the interaction between environmental parameters and virulence. MANAGEMENT IMPLICATIONS The existence of P . marinus races that vary in virulence or environmental tolerance has important management implications. Management programs should be sensitized to the potential dan- gers of spreading P . marinus races when relaying oysters, restock- ing oyster beds, or regulating effluents from shucking houses. Certainly, spreading virulent races should be avoided. Spreading races with varying environmental tolerances can be equally harm- ful by producing epizootics in areas that are inhospitable to indig- enous parasite races. Perhaps the most important implication of parasite races per- tains to the development of resistant oyster stocks. Regardless of the strategy employed (i.e. traditional breeding, chromosome set manipulation, hybridization with resistant species, and introduc- tion of resistant species), the existence of races complicates the evaluation of resistant stocks. Will a resistant stock developed against one race be resistant to other races? The lack of a race- specific interaction between P . marinus and C. virginica (Bushek and Allen 1996) indicates that the answer is yes. It is worth noting that the "resistant"" Texas population in Bushek and Allen's (1996) study continues to experience epizootic mortalities (Craig et al. 1989) despite nearly 50 years of natural selection. The population genetic structure of P . marinus is currently unknown but should be a high priority for basic and applied re- search, particularly with respect to management of oyster popula- tions. At this point, we do not know whether researchers currently studying P. marimis in different locales are working with similar or distinct races. Their results may not be comparable or transfer- able. An even worse possibility is that in vitro isolates may rep- resent single clones that happen to be well suited to in vitro pro- liferation and therefore may represent a small fraction of the ge- netic variation present in wild populations. Understanding the population genetic structure of P . marinus will help lead to a better understanding of the mechanisms that enable P . marinus races to spread. For example, combined with molecular tools, knowledge of the population genetic structure will help discriminate among the hypotheses commonly offered to explain the recent northward migration of P. marinus (Ford 1992. 1996). A continuous cline of genotypic frequencies in the popu- lation genetic structure from Chesapeake Bay northward would implicate natural range expansion, possibly permitted by recent regional climate warming. Alternatively, association of P. mari- nus isolates from the south with new epizootics in the north would imply that importation of southern oysters to northern habitats is responsible for the range expansion. Such an implication has been made regarding the origin of Haplosporidium nelsoni. which causes MSX disease in oysters on the East Coast of the United States. Recent studies (Friedman and Hedrick 1995) have de- scribed a Haplosporidian parasite in Pacific oysters (Crassostrea gigas) from Japan and California that is morphologically similar to H. nelsoni. Using a recently developed molecular probe for H. nelsoni (Stokes and Burreson. 1995). Stokes. Burreson. and Fried- man (unpublished data) demonstrated genetic identity among par- asites from each area. The presumption is that H. nelsoni was introduced to both coasts of the United States via the importation of infected oysters from Japan. Population genetic data on P. marinus will help identify the geographic source of those now in the northeast. Finally, the evolution of new strains of P . marinus with, for example, increased tolerance to cold temperatures or low salinity may be implicated by population genetic data if northern isolates are genetically unique and exhibit enhanced environmental tolerances compared to other isolates. These hypotheses are not mutually exclusive, but lacking information on P. marimis popu- lation structure, any are difficult to rule out. Despite nearly 50 years of research on P. marinus, the studies by Perkins and Menzel (1966) and Bushek and Allen (1996) are the only investigations that have attempted to compare isolates of P. marinus. Although Bushek and Allen (1996) demonstrated that races of P. marinus vary in virulence, confirmation of these results and the intricacies of genetic population structure (e.g.. are there more than two races?) await the precision of molecular tools and additional studies employing larger sample sizes. Once these have been developed, managers should be able to use molecular mark- ers as a forensic tool to identify source populations of epizootics and as a preventative tool to stop the spread of virulent races. ACKNOWLEDGMENTS We thank Pat Gaffney and an anonymous reviewer for provid- ing valuable comments on the manuscript and Bill Fisher for in- viting us to write this review. This is contribution number 1045 of the Belle W. Baruch Institute for Marine Biology and Coastal Research. LITERATURE CITED Andrews, J. D. 1988. Epizootiology of the disease caused by the pathogen Perkinsus marinus and its effects on the oyster industry. Amer. Fish. Soc. Spec. Publ. 18:47-63. Andrews, J. D. 1996. History of Perkinsus marinus. a pathogen of oysters in Chesapeake Bay 1950-1984. J. Shellfish Res. 15:13-16. Andrews, J. D. & W. G. Hewalt. 1957. Oyster mortality studies in Vir- ginia II. The fungus disease caused by Dermocystidmm mannuni in oysters of Chesapeake Bay. Ecol. Monogr. 27:1-26. Azevedo, C. 1989. Fine structure of Perkinsus ailanlicus n. sp. (Apicom- plexa, Perkinsea) parasite of the clam Rudilupes decussatus from Por- tugal. J. Parasilol. 75:627-635. Barber, B. J.. S. E. Ford & R, N. Wargo. 1991. Genetic variation in the timing and gonadal maturation and spawning of the eastern oyster. Crassostrea virginica (Gmelin). Biol. Bull. 181:216-221. Burreson. E. M. & J. D. Andrews. 1988. Unusual intensification of Ches- apeake Bay oyster diseases during recent drought conditions. Proc. Oceans 88: pp. 789-802. Burreson. E. M., R. S. Alverez. V. V. Martinez & L. A. Macedo. 1994. Perkinsus marinus (Apiconiplexa) as a potential source of oyster Cras- sosirea virginica mortality in coastal lagoons of Tabasco. Mexico. Dis. Aqiial. Org. 20(l):77-82. Bushek, D. 1994. Dermo disease in American oysters: Genetics of host- Races of P. m.\rinus 107 parasite interactions. Ph.D. dissertation. Rutgers-The State Univer- sity. New Brunswick. New Jersey. 189 pp. Bushek. D. & S. K. Allen. Jr. 1996. Host-parasite interactions among broadly distributed populations of the eastern oyster Craisosirea \ir- ginica and the protozoan Pfrkin.sus mariniis . Mar. Ecol. Progr. Ser. (In press). Bushek, D.. S. E. Ford & S. K. Allen. Jr. 1994. Evaluation of methods using Ray's fluid thioglycollate medium for diagnosis of Perkiiuus marinus infection in the eastern oyster. Crassoslrea virgumci. Anmi- Rev. Fish. Dis 4:201-217. Craig. A.. E. N. Powell. R. R. Fay&J. M. Brooks. 1989. Distribution of Perkinsiis mannii.s in gulf coast oyster populations. E.stiuiries 12(2): 82-91. Day. P. R 1974. Genelics i>t Hosl-Pcirasiw InliiLiction. W, H. Freeman and Co.. San Francisco, California, 238 pp. Dungan, C. F. & B. S. Roberson. 1993. Flow cytometric quantification and analysis of Perkinsiis marinus cells present in estuarine waters. Completion Report. NOAA NMFS Oyster Disease Research Program, contr. no. NA16FL0406-01, 23 pp. Faisal. M., J F. La Peyre & D Y Schafhauser. 1994. Detection of proteases in the supemates of Perkinsiis marinus cultures. J Shellfish Res. (abstract) 13(1):296. Ford, S. E. 1992. Avoiding the transmission of disease in commercial culture of molluscs, with special reference to Perkinsus marinus (Demio) and Haplosporuliiim nelsoni (MSX). J. Shellfish Res. 1 l(2l: 539-546. Ford, S. E. 1996. Range extension by the oyster parasite P. marinus into the northeastern United States: Response to climate change? J. Shell- fish Res. 15:45-56. Friedman. C. S. & R. P. Hedrick. 1995 Haplosporidian infections of the Pacific oyster, Crassoslrea gigas. Thunberg. J. Shellfish Res. (ab- stract) I4(l):266. Gauthier. J. D. & G. R. Vasta. 1993. Continuous in xiiro culture of the eastern oyster parasite Perkinsus marinus. J Imeriehr Pathol 62: 321-323. Goggin. C. L. & R. J. G. Lester. 1987. Occurrence of Perkinsus marinus species (Protozoa, Apicomplexa) in bivalves from the Great Bamer Reef. Dis. Aquat. Org. 3:113-117 King. R. C. & W, D. Stansfield. 1990. A Dictionary of Genetics. 4th Edition. Oxford University Press. New York, 406 pp. Kleinschuster. S. J. & S. L, Swink. 1993. A simple method for the in vitro culture of Perkinsus marinus. Nautilus 107(2):76-78- Krebs, C. J. 1985. Ecology: The Experimental Analysis of Distribution and Abundance. 3rd edition. Harper and Row Publ., New York, 800 PP La Peyre, J. F.. M. Faisal & E. M. Burreson. 1993. In vitro propagation of the protozoan Perkinsus marinus. a pathogen of the eastern oyster. Crassoslrea virginica. J. Euk. Microbiol. 40:304-310. Levine. N. D. 1988. The Protozoan Phylum Apicomplexa Vol I CRC Press. Inc.. Boca Raton. Florida. 144 pp. Mackin. J. G. 1956. Dermocyslidium marinum and salinity. Proc. Natl. Shellfish. Assoc. 46:116-12^ Mackin. J. G. 1962. Oyster disease caused by Dermocyslidium marinum and other microorganisms in Louisiana. Publ. Inst. Mar. Set. 7:132- 299. Mackin. J. G. H.. M. Owen & A. Collier. 1950. Preliminary note on the occurrence of a new protojoan parasite. Dermocystidium marinum n. sp. in Crassoslrea virginica (Gmelin). Science 111:328-329. Nelson. R. R. 1973. Breeding Plaius for Disease Resistance. The Penn- sylvania State University Press. University Park. Pennsylvania. 401 pp. NOAA Chesapeake Bay Office. 1995 Summaries of current ODR pro- jects. NOAA Chesapeake Bay Office. Suite 107-A 410 Severn Ave.. Annapolis. Maryland. Photocopy. 16 pp. Perkins. F. O. & R, W. Menzel. 1966. Morphological and cultural studies of a motile stage in the life cycle of Dermocystidium marinum. Proc. Natl. Shellfish. Assoc. 56:23-30. Ragone. L. M. & E. M. Burreson. 1993. Effects of salinity on infection progression and pathogenicity of Perkinsus marinus in the eastern oys- ter. Crassoslrea virginica (Gmelin). J. Shellfish Res. 12(l):l-7. Ray. S. M. 1954. Biological Studies of Dermocystidium marinum. The Rice Institute Pamphlet. Special Issue. Ray, S. M. 1966. A review of the culture method for detecting Dermo- cystidium marinum with suggested modifications. Proc. Natl. Shellfish Assoc. 54:55-69. Ray, S. M. 1996. Historical perspective on Perkinsus marinus disease of oysters in the Gulf of Mexico. J. Shellfish Res. 15:9-1 1 . Soniat. T. M. 1985. Changes in levels of infection of oysters by Perkinsus marinus. with special reference to the interaction of temperature and salinity upon parasitism. Northeast Gulf Sci. 7(2):I71-174. Soniat. T. M. & J. D. Gauthier. 1989. The prevalence and intensity of P. marinus from the mid northern Gulf of Mexico, with comments on the relationship of the oyster parasite to temperature and salinity. Tulane Stud. ZoolBot. 27(l):21-27. Stauber. L. A. 1950. The problem of physiological species with special reference to oysters and oyster drills. Ecology 31( 1 1:109-1 18. Stokes. N. & E. M. Burreson. 1995. A sensitive and specific DNA probe for the oyster pathogen Haplosporidium nelsoni. J. Shellfish Res. (ab- stract) 14(1):279. White. M. E.. E. N. Powell. S. M. Ray & E. A. Wilson, 1987. Host-to- host transmission of Perkinsus marinus in oyster {Crassoslrea virgin- ica) populations by the ectoparasitic snail Boonea impressa (Pyra- midelhdae). J. Shellfish Res. 6(1): 1-5. Wilson. A. W., E. N. Powell. M. A. Craig. T. L. Wade & J. M. Brooks. 1990. The distribution of Perkinsus marinus in Gulf coast oysters: Its relationship with temperature, reproduction, and pollutant body bur- den. //)/. Rev. Hydrobiol. 75(4):533-550, Joiinial of Sheltfish Research. Vol 15. No, 1, 109-117. 19%, A WHOLE-OYSTER PROCEDURE FOR DIAGNOSIS OF PERKINSUS MARINUS DISEASE USING RAY'S FLUID THIOGLYCOLLATE CULTURE MEDIUM WILLIAM S. FISHER AND LEAH M. OLIVER U.S. Environmental Protection Agency Center for Marine and Estiiarinc Di.sea.sc Research Gulf Ecology Division National Health and Environmental Effects Research Laboratory Gulf Breeze. Florida 32561 ABSTRACT Diagnosis of Perkin.sii.s mariniis disease of eastern oysters Crassostreii virginicu has been routinely accomplished by incubating oyster tissues in a fluid thioglycollate medium described by Ray in the early 1950s. At least three modifications of the technique are available with applications to different diagnostic needs. Of these, the quantitative whole-oyster technique is potentially the most valuable because it includes all oyster tissues and does not rely on subjective estimates of intensity. A variety of protocols and approaches were examined in an attempt to develop a standardized procedure for quantitative whole-oyster diagnosis that optimizes sensitivity, specificity, precision and accuracy. A recommended procedure, with possible variations, is presented here with the expectation that its presentation will foster further refinement and improvement. We conclude that the recommended whole-oyster diagnostic technique is capable of providing reliable quantifications of prevalence and intensity and has great potential for examining correlations of total P. marimis body burdens with measurements of oyster biology and for evaluating or calibrating other diagnostic techniques. A'£>' WORDS: Oyster diseases. Crassostrea virginica. Perkiii.siis rminiiKS. disease diagnosis INTRODUCTION The discovery of Perkinsus moriiuis ( = Dennocysndiiim inari- num) as a pathogen of eastern oysters (Crassostrea viri;inica) was made from oysters collected in the Gulf of Mexico through exam- ination of histological sections (Mackin et al. 1950). Soon after, Ray (1952a. 1952b) described a new technique that reduced the time, labor and equipment required for diagnosis and was more sensitive to detection of light infections. The technique was based on incubation of oyster tissue in (Ray's modified) fluid thioglycol- late medium (RFTM) to enlarge the parasites for greater visibility. The standard fluid thioglycollate medium was modified by rehy- dration with sea water rather than fresh water and by addition of antibiotics and antimycotics to reduce bacterial and fungal con- tamination. Samples from infected oyster tissues formed well- developed and enlarged parasites that were "conspicuous and readily identifiable" (Ray 1952a) after 10-18 hr incubation in RFTM. Parasites were visible in whole mounts (tissue squashes) and recognized by their thin-walled, spherical, cyst-like bodies measuring up to 35 jxm in size. After 72 hr incubation the size increased to 90 [im and could reach 150 [im after a week. As they enlarged, the parasites were observed to have thicker cell walls, increased vacuolization and greater lipid drop accumulation (Ray 1952a). Detection of infections, including light infections with smaller, less-developed parasites, was enhanced by staining with a [25-fold] aqueous dilution of Lugol's iodine. The cell walls of parasites incubated In RFTM for 18 hr. 36 hr and a week were progressively stained faint blue, a distinct dark blue, and blue- black. The ability of detect light infections using iodine stain was dramatically improved over unstained tissue squashes. Early evidence Indicated that parasite numbers did not increase during incubation In RFTM. even after several months (Ray 1952b), and further studies corroborated these findings (Stein and Mackin 1957). Since parasites evidently did not multiply in cul- ture, the RFTM technique could be used to estimate Infection intensity as well as prevalence. Thus, a relative intensity scale was established to delineate light, moderate and heavy Infections (Ray et al. 1953). A rating system, devised by Mackin and first de- scribed by Ray ( 1954a), assigned values to six subjective intensity categories, then divided the total by the number of diagnosed Individuals to obtain a "weighted incidence," I.e.. an average Infection Intensity of the population sampled.' This method of determining average infection Intensity, further characterized by Mackin ( 1962), has been used widely to describe the epizootiology of P. marinus disease (Andrews 1988). Although semi- quantitative, the rating system has also been successfully used to profile disease progression, to demonstrate the strong relationship of high-intensity infections with mortalities and to correlate infec- tion Intensity with physiological changes (Soniat and Koenig 1982). Lewis et al. (1988) and Choi et al. (1989) provided an Impor- tant alternative for the RFTM procedure, A major obstacle to enumerating parasites had been the obstruction of parasites by oyster tissue, i.e.. they were not sufficiently isolated for quanti- fication. These researchers used NaOH to digest oyster tissues without damaging the integrity of the prezoosporangia cell wall. Choi et al. (1989) used the method on dissected tissues to dem- onstrate that the commonly used Mackin scale was exponential. They could then retrospectively use Mackin scale ratings as an estimate of total parasite numbers. The digestion of tissues led Gauthier and Fisher ( 1990) and Bushek et al. (1994) to introduce two new quantitative assays using RFTM. Consequently, there are now three modifications of RFTM cul- ture to characterize P. marinus infection intensity in Individual oysters; ( 1 ) an exponential estimate from selected tissues using the semi-quantltativc Mackin ratings (Ray 1954a, Mackin 1962), (2) a quantitation of parasite density in oyster hemolymph ( Gauthier and 'Although "incidence" was once used interchangeably with "preva- lence," the current definition for incidence is the rate of occurrence of new cases of a particular infection in a population (Overstreet 1978, Margolis et al. 1982) "Weighted prevalence." as is currently used, is a more acceptable term for this calculated value. 109 110 Fisher and Oliver Fisher 1990). and (3) a quantitative whole-oyster diagnosis (Bushek et al. 1994). The value of these different techniques de- pends on their application (Bushek et al. 1994). For example, the tissue squash technique has proven invaluable in epizootiological studies (Andrews 1988. 1996, Burreson and Ragone Calvo 1996. Ford 1996. Soniat 1996) and the hemolymph biopsy technique can be used when animals cannot be sacrificed (Fisher et al. 1992). Yet the whole-oyster technique is potentially the most valuable because it is quantitative and incorporates all oyster tissues. These attributes make it the standard by which other techniques should be evaluated and the most suitable technique for correlation of disease intensity with measurements of oyster biology. Researchers are now seeking to relate infection intensity to measurements of oyster physiology, immunology, reproduction and growth, in attempts to understand susceptibility of oysters and sublethal effects of P . marinus disease (Anderson 1996. Chu 1996. Paynter 1996). The most supportable correlations will stem from reliable whole-oyster quantifications of disease prevalence and intensity. The value of any diagnostic technique depends on attaining certain standards of sensitivity, specificity, precision and accuracy (Bushek et al. 1994). These standards must be continually bal- anced with cost. ease, technical capability and purpose. The fol- lowing section reviews some inherent assumptions when using RFTM methodology for diagnosis of P. mannus and the potential impact of those assumptions on diagnostic standards. ASSUMPTIONS OF RFTM DIAGNOSIS All diagnostic tests are encumbered with certain assumptions. the validity of which will affect the reliability of the test (Bushek et al. 1994). Assumptions of RFTM methods for diagnosis of P . marinus disease of eastern oysters involve at least the following issues: retrieval of parasites from the host tissue; detection: quan- tification: stability of parasite numbers during processing: and the representativeness for the tissue, organism or population sampled. Retrieval of Parasites From Host Tissue One assumption of the RFTM diagnostic test is that all para- sites, including different life stages, are retrieved from the host. If oysters are being examined for possible introduction into a dis- ease-free area, this becomes a precarious assumption; any false- negative diagnoses could introduce the parasite into a naive or uninfected oyster population. Single-tissue techniques (such as the tissue squash and hemolymph diagnoses) cannot adequately ensure the absence of parasites in other tissues. Obviously, diagnosis of the whole oyster avoids loss of parasites from unsampled tissue. The possibility also exists that preparation or processing of oyster tissue for diagnosis destroys or masks certain parasite life stages. This possibility was difficult to test when alternative meth- ods of diagnosis were lacking. Until recently, only histological examination of paraffin sections was available for comparison with the RFTM technique: Ray (1954b) and Stein and Mackin (1957) provided evidence that all known or identifiable stages of P. manmis did enlarge in RFTM. Although reassuring, these stud- ies lacked a strong comparable method to disprove the potential presence of unidentified life stages. Recent development of an immunoassay technique (Dungan and Roberson 1993) has provided an alternative means to detect P. marinus. Moreover, the technique is founded on molecular rec- ognition that should identify any stage off. marinus that does not present unique epitopes. Using the immunoassay technique. Ragone Calvo and Burreson ( 1994) did not detect previously un- described cryptic stages of P. marinus in winter samples of oysters although they were able to detect light infections in oysters diag- nosed as negative by RFTM culture of tissue and hemolymph. If further research verifies these findings, then there will be reason- able assurance that RFTM culture techniques retrieve and enlarge all forms of P. marinus. Detection of Parasites It must also be assumed for any RFTM test that all P. marinus retrieved from the host are detectable by the procedures employed. Techniques using RFTM have two significant attributes that en- hance recognition and quantification: enlargement of the parasite and enhanced uptake of iodine stain. When enlarged, the parasites are distinguishable by morphological characteristics (double cell wall, signet ring appearance) (Perkins 1988. 1993, 1996); when enlarged and stained with iodine, the parasites become so easily recognized that quantification becomes feasible. Enlargement is an important attribute of any RFTM technique. Ray (1952a, 1952b) noted that increasing numbers of parasites were detectable in RRM culture over the first 1 8 hr of incubation, but not thereafter. This was interpreted to mean that it took 18 hr incubation for all parasites present to become sufficiently large for visual detection. Consequently, he recommended that a minimum incubation period of 48 hr would ensure enlargement of all para- sites to a detectable size. However, after surprisingly low counts were recorded in field studies (Ray 1966a). he suggested that the incubation time be increased to I wk (Ray 1966b). Ray (1966b) also changed antibiotic components of RFTM based on the need to enlarge the parasites. In studies comparing different antibiotic regimes, he found that penicillin, streptomycin and Chloromycetin inhibited parasite enlargement, but this effect was ""spared" by addition of the antifungal agent mycostatin. Trials showed infection intensities as much as 25 times higher with mycostatin in the formulation. The lower counts (without mycos- tatin) were believed to be due to a lack of parasite enlargement, since higher magnification revealed small, otherwise unidentifi- able cells in the preparation. Thus, mycostatin was included in the RFTM formulation, not only as a useful broad-spectrum antifungal agent, but because it allowed the parasites to enlarge in the pres- ence of other antibiotics. In a number of recent studies, mycostatin was inexplicably deleted from the formulation; it is unknown whether this practice affects the number of prezoosporangia de- tected. Enlargement of P. marinus is due to their uptake of RFTM. and uptake occurs only if the parasites are living. Thus, accidental diagnosis of dead parasites is unlikely since they would remain too small for detection. It is presumed that all living parasites placed in RFTM survive the incubation period, although this has not been examined experimentally. Techniques that employ NaOH to dis- solve oyster tissue (and. perhaps, kill parasites) do not impact en- largement because RFTM incubation precedes NaOH digestion. Staining is particularly important for research efforts that at- tempt to quantify or estimate infection intensities. Ray (1952b) stated that the blue color of parasites after exposure to Lugol's iodine was not due to reaction with starch, but rather to a gluco- side. such as saponarin. in the cell wall (using fungal terminol- ogy). Even with this diagnostic aid. quantitative precision and accuracy are not guaranteed. Since staining with Lugol's iodine requires enlargement in RFTM (Mackin 1962). unstained parasites Whole-Oyster Diagnosis of Perkinsus II I or those that are too small to be seen could cause artificially low counts. Artificially high counts may result if the iodine is too concentrated; parasite morphology becomes less distinguishable from stained debris and false-positives can result from precipitated iodine. The specificity of the stain for P. maritms is another important consideration. To be completely effective, the stain must bind with all forms of P . maninis and with no other parasite. There is rea- sonable evidence that all known life stages of P. maiinus can be stained with Lugol's iodine after enlargement in RFTM, but it is not clear whether P . marinus is the only parasite being stained. There has been concern over possible confusion with members of the family Thraustichitriidae. but these do not closely resemble Perkinsus and the stain appears brownish (F. Perkins, personal communication). There are some fish myxosporean parasites (Schmidt and Roberts 1977) that stain with iodine, but these have not been observed in bivalves. The greatest potential for confusion is with other species oi Perkinsus (Lester and Davis 19S1 , Goggin and Lester 1987, Perkins 1993). Certainly Perkinsus sp. (Perkin- sus atlwiticus?) normally found in Macoma balthica could be mistaken for P . marinus. and considering the results of cross- infection experiments performed by Goggin et al. ( 19891, it would not be extraordinary if they occurred in C . virgmicu. Enlargement and staining by the RFTM technique are specific for Perkinsus only at the genus level. The polyclonal antibody developed by Dungan and Roberson ( 1993) is also specific for Perkinsus only at the genus level; thus, confirmation at the species level must await development of techniques such as genetic probes (Marsh and Vasta 1995). Recent quantitative techniques have generated additional as- sumptions related to staining. The use of 2M NaOH to digest oyster tissue does not appear to affect staining of the parasites as long as samples are washed free of NaOH, but no studies have confirmed this. Basic conditions cause iodine staining to fade quickly (D. Bushek, personal communication), but again no stud- ies have compared counts of stained P marinus with and without treatment in NaOH. Overall, the use of NaOH should reduce er- rors from stained debris and should improve staining specificity since most nontarget parasites will not withstand the harsh NaOH treatment. Quantification of Parasites The semi-quantitative tissue squash has been used only as an estimate of infection intensity. Even so, the subjective nature of the rating scale predisposes the technique to researcher bias. Ray (1966b) recognized the potential problem; Since intensity rating is a subjective procedure . . . there may be a tendency to rate an infection higher for tissues with large {parasitel cells than tor tissues with a similar concentration of small cells (p. 64) Other variables, such as stain intensity and tissue thickness, can also influence the assigned ratings. Lewis et al. (1988) recognized that the primary obstacle to quantification of P. marinus was the inability to separate prezoo- sporangia from oyster tissues. They found that digestion of oyster tissue in 0.5% trypsin followed by 2M NaOH provided a prepa- ration of parasites free of oyster tissue and readily quantifiable. Choi et al. (1989) used only 2M NaOH, but accomplished the same objective as they were able to enumerate P marinus from individual tissues after digestion. Quantification after NaOH di- gestion has now been used for several studies (Gauthier and Fisher 1990, Fisher et al. 1992, 1995, Bushek et al. 1994). Tissue digestion in the protocol does not alleviate all of the problems associated with quantifying P. marinus. Some stained particles may still not be recognized as P. marimis because they are much smaller than neighboring particles. Even after digestion, P. marinus are not necessarily distributed uniformly in the prep- aration (Choi et al. 1989, Gauthier and Fisher 1990). Accurate counting is also a source of error, particularly since parasites are usually too numerous to quantify without diluting the sample. Prezoosporangia can adhere to each other, forming clumps and chains that will distort results even when the most careful dilutions are made. Stability of Parasite Numbers For any estimate or quantification of parasites, it is critical that the numbers of parasites do not increase or decrease after the sample is collected. The greatest concern has been that parasites increase in number during incubation in RFTM. Yet, Ray ( 1952b) found only four cases in several hundred where parasite numbers increased, and incubation times for these ranged from 23 d to 7 months. Parasite multiplication may have occurred prior to en- largement or may have been due to enlarged parasites that pro- duced budding hyphae (fungal terminology). In later studies, Ray ( 1954b) concluded that the number of all microscopically identi- fiable stages off. marinus could increase slightly during 8-18 hr incubation in RFTM. but not thereafter. Mackin and Boswell (1956) and Mackin (1962) found multi- plication of P . marinus in very dilute thioglycollate medium dur- ing studies to describe the saprophytic cycle of P. marinus. The parasite was found to multiply in dilutions ranging from [10;1] to 1100;11 RFTM, with the best multiplication found in a |50:1] dilution. No multiplication was found in full-strength RfTM. In cases where multiplication did occur, it was never resolved wheth- er the nutrient source was dilute RFTM or remnant pieces of oyster tissue that were added with the inoculation. Other attempts to culture P. maritms using full-strength RFTM have also generally failed (Ray 1954b. Mackin 1962, Prokop 1950, Mackin and Boswell 1956). Stein and Mackin (1957) monitored P. marinus closely to identify all life stages during RFTM incubation. They found that all known forms of the parasite were enlarged and that reproduction was minimal. Although no studies have specifically addressed the issue, dy- ing parasites that are unable to enlarge in RFTM could artificially lower the measured intensity. However, results from many studies showing relative stability of counts from the same sample over time indicate that parasite mortality during incubation is not a critical concern. Representation of Tissue. Organism, and Population One of the most challenging diagnostic issues is whether values obtained with RFTM diagnoses adequately reflect the true infec- tion intensity of a given tissue, oyster or population of oysters. Early histological studies (Mackin 1951) and RFTM diagnoses (Ray 1952a) illustrated the fact that P. marinus are not evenly distributed among different oyster tissues or even in different sec- tions of the same tissue. Consequently, techniques employing only one tissue, or one section of a tissue, could easily misrepresent the actual intensity. This is an important concern for both prevalence and intensity measurements. Fisher and Oliver Ray (1952b) recognized that different applications of the RFTM technique would require different approaches. He sug- gested that four tissues (gill, mantle, heart and rectum) should be examined for a thorough diagnosis of light infections, but allowed that only the rectum was necessary for survey work involving large numbers of oysters. In comparative studies. Ray (1966a) showed a slightly higher prevalence in rectal than in mantle tissue, but he favored using a section of the anterior mantle due to ease of dis- section, which resulted in less opportunity for contamination. He also noted (Ray 1966b) that use of rectal tissue could be a disad- vantage when oysters have a well-developed gonad, because of difficulty in separating tissues and interference from adductor muscle fragments. Choi et al. (1989) demonstrated differences in infection intensity among digestive gland, mantle and gill tissues from the same oysters. Bushek et al. (1994) compared rectal and mantle tissue squashes throughout an annual period and found essentially no differences in overall sensitivity, but combining results from both tissues lowered the likelihood of false-negative diagnoses. Ray ( 1966b) also concluded that using rectum, gill and mantle as a composite preparation probably gave a better indica- tion of infection than a single tissue. Some results (Ray 1966b) using the RFTM tissue squash tech- nique found light P. mahmis infections in oyster gills but not in rectal and mantle tissues. He suggested that at least some oysters may be invaded by way of the gills rather than the digestive epi- thelia as indicated by the histological studies of Mackin (1951). in more advanced infections, mantle and rectal tissues had higher parasite ratings than did gill tissues. It is likely that localization of parasites in tissues may be partly dependent on the route of infec- tion and subsequent progression of the disease. Gauthier and Fisher (1990) employed RFTM to diagnose P. mariniis from oyster hemolymph samples. The principal purpose of using hemolymph was the ability to enumerate parasites without sacrificing the animal, thereby providing a means to study pro- gression of disease in living animals. Results indicated that the density of parasites in 1-mL hemolymph samples was comparable to the subjective ratings of mantle tissue from the same individuals with the added advantage of detecting many light infections that were negative with the mantle tissue squash (Gauthier and Fisher 1990). However. Bushek et al. (1994). using 250-|jlL hemolymph samples, found no difference in sensitivity between hemolymph diagnosis and the combined ratings of mantle and rectal tissue squashes. These authors noted that differences between the two studies may have been due to the lower hemolymph volume ana- lyzed and the combined ratings of rectal and mantle tissue used in their study (Bushek et al. 1994). Uneven parasite distribution in tissues is probably the strongest argument for development of a whole-oyster diagnostic technique. Quantification of whole oysters would not have been possible without the procedures of Lewis et al. (1988) and Choi et al. (1989) to extract P. marinus prezoosporangia from obstructing oyster tissues. Capitalizing on this technique. Bushek et al. (1994) published a quantitative whole-oyster diagnostic protocol in their evaluation of tissue squash and hemolymph diagnosis. The whole- oyster technique served well as a baseline for the comparison because, as expected, it was the most sensitive of the three pro- tocols. Separately, a quantitative whole-oyster protocol was gen- erated at the Environmental Protection Agency's Gulf Ecology Division (EPA/GED) to assess the effects of tributyltin on disease susceptibility (Fisher et al. 1995) and to examine annual intensity fluctuations of P. marinus in three bays in North America (Oliver et al. 1996). The basic principles of the Bushek et al. (1994) and EPA/GED protocols were the same, but several technical details varied. These details were considered in the following develop- ment of a protocol that should, at least temporarily, standardize the approach and serve as a template for further refinement. QUANTITATIVE RFTM DIAGNOSIS USING WHOLE OYSTERS The benefits of a quantitative whole-oyster diagnostic proce- dure may not offset the time and cost of performance for most monitoring (epizootiological) studies that require average popula- tion intensity and prevalence estimates. It is. however, a technique that has great potential for correlating total body burdens with measurements of oyster biology and may be particularly useful as a standard for evaluating other RFTM techniques. Recognizing the potential of these and other applications for a quantitative whole- oyster diagnostic technique, wc have examined different proce- dural components to develop a protocol that would optimize sen- sitivity, specificity, precision and accuracy. The protocol is based on that of Lewis et al. (1988) and Choi et al. (1989) and incor- porates procedures of Bushek et al. ( 1994) and those developed at EPA/GED. A similar protocol was appraised and performed by E. Burreson and L. Ragone Calvo at Virginia Institute of Marine Science. College of William and Mary, and by S. Ford and J. Gandy at Haskin Shellfish Research Laboratory. Rutgers-The State University. Some of their results are reported here. Also, valuable recommendations have been made by Drs. Burreson. Ford and D. Bushek (Belle W. Baruch Institute for Marine Biol- ogy and Coastal Research. University of South Carolina) during the preparation of this text. A summary of the recommended protocol is provided (Table 1). Minced or homogenized oyster tissues are placed into RFTM for enlargement of prezoosporangia. After RFTM incubation, preparations are placed into 2M NaOH, which digests oyster tissue but does not destroy parasites. After centrifugation and washing, the extracted parasites are stained with iodine, diluted, and aspi- rated onto a filter paper for counting at low magnification. Various aspects of this scheme require clarification and qualification, and some recommended steps have valid alternatives. These consid- erations are presented under the following sections on sample preparation. RFTM incubation. NaOH digestion, storage of ex- tracted parasites, iodine staining and parasite dilution, identifica- tion and counting. Sample Preparation Oysters should be kept cold after collection to reduce prolifer- ation of the parasite prior to processing. When oysters are shucked, care must be taken to retain all hemolymph with the sample to ensure that P. marinus associated with hemolymph is included. Oyster tissues should be diced or minced so that RFTM can easily penetrate tissues to uniformly reach all parasites. Also. finer pieces will enhance subsequent tissue digestion with NaOH. In most studies, sample replication is not performed until the end of the diagnostic process, i.e.. with replicate counts of aliquots taken from the same tube. In such a case, it is not necessary to homogenize oyster tissues into a slurry before incubation in RFTM. However, some studies, such as interlaboratory compar- isons, quality assurance exercises or evaluations of reproducibili- ty, may require sample replication earlier in the process. In such cases, sample homogeneity is critical and tissues should be mixed in a blender and tissue grinder. Whole-Oyster Diagnosis of Perkinsus 113 TABLE I. Summary of protocol for whole oyster dia)>nosis. Reagent preparation A RFTM Heat 242.5 niL of distilled water to a boil, stir in 7.3 g of thioglycollate and 5 g of NaCI Remove from heat and dispense into 50-mL culture tubes. Autoclave for 15 min at 15-lb pressure and 250°C (color should be orange/brown). Store in the dark al room temperature. B. Antibiotics Mix 0.25 g Chloromycetin (Sigma No. C-0378) with lOO niL of distilled water, shake well and refrigerate. Mix 0.01 g mycostatin or nystatin (Sigma No. N--?503. 5160 USP units/mg) with 10 niL of distilled water, shake well and refrigerate . C Lugol's iodine Stock solution: Mix 6.0 g of potassium iodide with 4() g of iodine and add to 100 mL of distilled water. Working solution = 1 niL of Lugol's stock solution mixed in 25 mL of distilled water. Tissue sample preparation and incubation A. Shuck the oyster carefully and obtain a wet weight. Dice the tissues into 2-5 mm sections and. if desired, homogenize in a blender and/or tissue grinder. B Place tissue homogenates into 50-niL tubes, add 20 mL of Rl-TM and 100 p.L of Chloromycetin and mix. then gently layer 2 niL of nystatin onto media. Do not mix Incubate tubes in the dark for 7 d at room temperature. C. Centrifuge at 1,500 x g for 10 min, aspirate and discard the RFTM supemate. Add 20 mL of 2M NaOH to the tubes and incubate m a 60°C water bath or oven for 2-6 hr, depending on tissue degradation. Centrifuge at 1,500 x g for 10 min, aspirate to remove NaOH supemate. Wash 3 times with 10 mL of deionized water, mixing well and centrifuging at 1,500 x g for 10 min; washed samples may be stored in refrigeration in 0.2'7t NaN,. D. Resuspend pelleted parasites in 1 niL of Lugol's |25:1| working iodine solution. Mix vigorously, place 100-(j.L aliquot on 0 22 jini-pore-size filter paper and aspirate. Use an ocular grid and 100 ■- magnification to count the dark blue parasites: If <20 prezoosporangia are found in 100 (xL, then count the entire (1-mL) sample. If >200 prezoosporangia are found in 100 jjlL, serially dilute by transfemng at least 100 piL of sample into less than 10-fold dilutions Multiply the means of three 100-|a.L aliquots by the appropriate dilution factor and record infection intensity as log,u prezoosporangia g" ' wet weight tissue. In one such exercise performed at EPA/GED, a heavily in- fected oyster was placed into a blender and homogenized for 30- 45 sec. The resulting chopped tissue was further processed using a tissue grinder with 0.1-mm clearance, producing a fine slurry. This was subdivided into 10 equal volumes and added to RFTM in 10 separate tubes for a l-wk incubation period. After incubation, a technician processed five subsamples through NaOH digestion (2 hr). staining and counting. The remaining five subsamples were identically processed by a different technician. Results of these counts exhibited a significant difference (Student's t-test. p < 0.05) between the means obtained by each technician (log,,, 7.05 vs. log,,, 7.26. range = logm 6.86-log,„ 7.36). The significant difference in this limited exercise illustrates the potential bias at- tributable to different individuals performing the diagnostic test, even though the same protocol, equipment and reagents were used. If possible, it appears highly desirable that sample process- ing and counting for a given study be conducted by the same technician. Sample processing and counting variability was examined through efforts with researchers at Virginia Institute of Marine Science and Haskin Shellfish Laboratory. At EPA/GED, slurries of four individual oysters were prepared as described and duplicate 1 .0-mL aliquots were dispensed into sterile test tubes with thor- ough mixing between each aliquot removal. Each tube was la- belled with a blind code, placed on ice in a cooler and shipped overnight to the collaborating laboratories. Blind coded samples that remained at EPA/GED were refrigerated overnight. Protocols for RFTM incubation, NaOH digestion and counting of prezoo- sporangia were performed by each laboratory according to a writ- ten protocol similar to that described in Table 1 . Several compar- isons were performed during 1994. In one interlaboratory comparison, good agreement in infection intensity was attained (Table 2a), as reflected by coefficients of variation below 10%. These data clearly support the potential reproducibility of the diagnostic procedure among researchers at different laboratories with different equipment and reagents, at least for samples containing moderate to high levels of infection. Another comparison (Table 2b) demonstrated an apparent weak- ness of the RFTM technique that may not be overcome by tech- nical improvements. Quantification of infection intensity was rea- sonably reproducible for oysters with moderate infection intensi- TABLE 2, Total number (log,„) of prezoosporangia per sample counted during two interlaboratory comparisons, labelled (al and (b), of oyster tissues infected with P. mahniis. Four oysters (1—4) were tested in each com- parison with duplicate samples of homogenate analyzed by each lab- oratory. The values recorded for each duplicate are the averages of three counts made from 100-p.L aliquots. Overall means were calcu- lated from counts of all six duplicates. Coefficients of variation (CV"7f ) were calculated as standard deviation divided by the mean times 100, Oyster Laboratory Mean No. A B C CV% (a) 1 3.14 3.24 3.75 2.91 3.44 3.72 3.37 9.9 2 5.24 5.31 5.92 5.62 5.56 5.62 5.52 4.5 3 5.62 5.56 5.84 5.26 5.48 5.58 5.56 3.4 4 3.80 3.07 3., 30 3.38 3.13 3.63 3.39 8.4 (b) 1 0.90 1.90 1.48 0.00 2.10 0.78 1.19 65.9 2 3.72 3.61 3.95 3.62 3.73 4.01 3.78 4.5 3 3.56 3.27 1.30 2.10 3.33 3.72 2.88 33.4 4 0.00 1.58 0.85 0.00 1.97 0.85 0.87 92.0 114 Fisher and Oliver ties (oysters #2 and #3) but highly variable for those with low infection intensities (oysters #1 and #4). This weakness was previously noted by Bushek et al. (1994) who found both he- molymph and tissue assays to lose sensitivity below 10"* parasites g ~ ' wet tissue weight. In our comparisons of whole body burdens, replication within each laboratory (i.e., between duplicate sam- ples) was not particularly poor at low intensities (Table 2bl. but coefficients of variation were high among laboratories. This pat- tern may indicate that to accurately quantify low-intensity infec- tions, other protocols, such as specific antibody (Dungan and Rob- erson 1993) or DNA probe (Marsh and Vasta 1995) techniques, may be required. RFTM Incubation P. manints prezoosporangia progressively enlarge from around 10 \i.m diameter (Mackin et al. 1950. Ray 1954b) to 35, 90 and up to 150 p-m diameter after 18 h. 72 h and 1 wk incubation in RFTM (Ray 1952a). If insufficient time or medium is available for all parasites to enlarge to a detectable size, lower counts will result. This protocol recommends that oyster tissues be placed into 50-mL tubes containing 20 niL RFTM (Table I ). The constant volume in all tubes allows RFTM to be prepared in advance and should not create discrepancies unless there is insufficient medium for all parasites to enlarge to a detectable level. However. Bushek et al. (1994). who placed 25 niL RFTM into each lube, found the size of P . marimis prezoosporangia in whole-oyster diagnoses inversely correlated to infection intensity. Although it was not investigated, parasite enlargement in high-intensity samples may have been cur- tailed by competition for limited resources in the medium, if so. then we must question whether enlargement can be so inhibited that some parasites will be too small for detection. To address this question, experiments must be performed that resolve the discrep- ancies of sample volume : RFTM volume noted by Bushek et al. (1994). Although other factors, such as biological cycles of par- asites related to their seasonal intensity, could create the small size of prezoosporangia in high-intensity samples, it should be consid- ered that greater volumes of RFTM may be needed. If so, the ratios of antibiotics added to RFTM in the culture tubes must be maintained. Ray ( 1952a. 1952b) originally suggested that a mminium 48-hr RFTM incubation period would ensure enlargement of all parasites present to a detectable size, but later altered this opinion to 1 wk (Ray 1966b). Most studies have reported incubation times from 5 to 7 d, but Bushek et al. ( 1994) used a 2-wk incubation period for whole-oyster diagnosis. A longer incubation period may serve to diminish concerns over insufficient medium. Centrifugation of the sample after RFTM incubation must re- liably pellet all of the parasites before removal of the supernate. For hcmolymph samples. Bushek et al. ( 1994) found no parasites remaining in RFTM supernate after centrifugation at 1 .000 x g for 20 min. The EPA/GED protocol for whole oysters recommended 1.500 X g for 10 min. although no studies were completed to evaluate speeds and durations. As an introductory experiment, we examined duplicate RFTM supernates of whole-oyster homoge- nates after centrifugation speeds of 280, 570, 1,1 18, 1,460 and 2,282 X g for 10 min. Supernates were counted after aspiration through a filter paper and iodine staining. Parasite intensity over all samples averaged 1 . 1 x 10* parasites g~ ' oyster tissue (s.d. = 0.08 X 10'') and percentages of parasites remaining in the super- nate for increasing speeds were, respectively. 0.011. 0.015. 0.008. 0.004 and 0.0049!:. Obviously, centrifugation at speeds as high as 2.282 x g for 10 min did not retrieve all of the parasites from RFTM. but at the density tested, the number of parasites lost would not appreciably affect the count (0.004%). The importance of this finding at other parasite densities is not yet known. High centrifugation speeds can pellet the parasites so tightly that they are hard to disperse for subsequent processing. Perhaps a longer duration would be the best procedure if these minor losses are unacceptable to the investigator. To avoid loss of pellet material, supernates should be removed by gentle aspiration rather than pouring. NaOH Treatment The whole-oyster technique was made possible when research- ers at Texas A&M University (Lewis et al. 1988. Choi et al. 1989) described a "hypnospore separation method" that dissolved oyster tissues from P. mariims prezoosporangia. Tissue digestion re- moved oyster debris to allow easier identification and counting of parasites. Choi et al. ( 1989) digested oyster tissues for 1 hr at 50°C in 20 mL 2M NaOH per gram of wet tissue weight and then centrifuged them at 1 .600 x g for 15 min. removed the supernate. and washed them four times in buffered saline before resuspending the parasites. The generation of this protocol was not detailed but has been used by others without apparent complication. Bushek et al. (1994) used only 10 niL 2M NaOH per gram of wet tissue weight. It was not clear from the published texts of Lewis et al. ( 1988) or Choi et al. ( 1989) whether NaOH digestion degraded or in any way altered the detectable number of parasites. This may have been difficult to examine in a controlled experiment because with- out digestion, enumeration of parasites from tissues can be highly variable. To shed some light on this question, an experiment was performed to compare parasite numbers after increasing durations of digestion. Both whole oysters and hemolymph were examined. Hcmolymph samples were included because there is little tissue debris in hemolymph that can interfere with counts in the control (no digestion) samples. In fact. Bushek etal. (1994) suggested that hcmolymph samples could be counted without NaOH digestion. Hemolymph (2 mL) from three oysters was withdrawn from the adductor muscle and placed in three individual tubes containing RFTM. The oysters were then shucked and tissues were finely homogenized and placed into three separate tubes containing RFTM. After incubation for 1 wk at room temperature, each he- molymph sample and each whole-oyster sample was divided into eight subsamples and centrifuged to remove RFTM. Duplicate subsamples were then digested in 10 mL 2M NaOH for 0. 2. 6 and 24 hr at 60°C. Following digestion, samples were washed and quantified according to the protocol (Table 1). Results of this experiment (Table 3) indicated no effect of NaOH on P. marimis counts. This was evidenced by the lack of change in the hemolymph samples and from the lack of diminution in the whole-oyster samples. As implied by Bushek et al. ( 1994). it does not appear from this experiment that digestion of oyster hemolymph is necessary for reasonable precision. However, di- gestion of whole-oyster tissues does appear necessary for quanti- tative applications. Undigested whole-oyster samples were diffi- cult to count because of interference by oyster tissues, and counts were generally not as high as digested samples. Also, longer di- gestion periods provided lower coefficients of variation in the whole-oyster samples, probably due to greater separation of pre- Whole-Oyster Diagnosis of Pt:KKiNsus 115 TABLE 3. Mean numbers of prezoosporangia (logn,l and coefficients of variation (C"V% I for three oysters analyzed for both tissue (g '» and hemolymph (mL ') levels. Means were derived from triplicate counts of duplicate samples. NaOH Digestion Period (hr) Oyster No. 0 24 Tissue samples 1 5.01(8%) 5.14(41%) 5.17(21%) 5.07(6%) 2 4.79(67%) 4.31(1%) 5.36(2%) 5.38(4%) 3 5.30(10%) 5.40(6%) 5.43(1%) 5.48(6%) Hemolymph .samples 1 ' 3.01(21%) 3.09(5%) 3.11(5%) 3.05(30%) 2 2.40(6%) 2.44(13%) 2.33(3%) 2.31(3%) 3 2.88(29%) 2.91(1%) 2.67(2%) 3.06(29%) zoosporangia from oyster tissues. For these reasons, a 2-6 hr digestion period is recommended in the protocol (Table 1 ), depen- dent on tissue degradation. Highly homogenized tissue samples need less digestion time. Even longer periods could be used if necessary: Although some artifacts were observed in samples di- gested for 24 hr. it does not appear that long digestion periods will reduce the parasite count. It also seems reasonable to redigest samples with remnant tissues without fear of altering the counts. Centrifugation and washing after NaOH digestion are impor- tant steps that must be completed with care. This may be more critical than centrifugation after RFTM incubation where parasites are less dense. If centrifugation is too light, then parasites are lost in the supernate. If too harsh, the parasites can form clumps that are difficult to disperse and interfere with counting. Also during NaOH digestion, parasites become sticky and may adhere to the walls of centrifuge tubes (D. Bushek. personal communication). Yet the parasites must be washed free of base or the iodine stain fades very quickly. Since available equipment varies among lab- oratories, it is important that researchers optimize post-RFTM and post-NaOH centrifugation speeds and times to balance these fac- tors. Storage of Extracted Parasites Once P. marinus prezoosporangia are extracted from oyster tissues and washed free of NaOH. they may be stored for several weeks before counting. The EPA/GED procedure resuspended parasites in 1 mL of 10 mg/mL NaCI and stored them in the refrigerator. Bushek et al. (1994) stored extracted parasites at room temperature in a phosphate-buffered solution containing 2% paraformaldehyde and 0.04% sodium azide (NaN,). Currently at the Haskin Shellfish Research Laboratory, parasites are resus- pended in 0.2% NaN, and samples are stored in the refrigerator (S. Ford, personal communication). Sodium azide is primarily used to reduce bacterial contamination but may also prevent clumping. Parasite clumping is a serious problem that can create high variability in replicate counts. Clumping is compounded by poor washing, high centrifugation speeds and prolonged storage. Most often, vortex mixing is employed to resuspend samples but is not always successful. It is possible that the addition of a soap could reduce the aggregation of hypnospores prior to staining. However, different concentrations of Tween 80 applied to prezoosporangia suspensions did not reduce their clumping and caused the iodine stain to tade more quickly. The possibility of using other soaps, such as Triton-X, should be investigated. Sonication for 1-2 min has been found to successfully disrupt clumps (S. Ford, personal communication) and is recommended for routine processing. In all cases, preparations should be examined before counting to deter- mine whether parasite clumping will cause misrepresentation of the sample. Iodine Staining As previously noted, incomplete digestion of oyster tissue could lead to obstruction of parasites during counting. Alterna- tively, remnant tissues that stain with iodine could artificially in- crease parasite counts. Ample digestion of oyster tissues is there- fore going to reduce problems associated with staining and count- ing of prezoosporangia. Even so. morphological identification of parasites is an important aspect of any RFTM diagnostic tech- nique. Proper staining with iodine greatly improves this capability, but overstaining can lead to poor identification. Ray (1952a) orig- inally recommended a |25:1] aqueous dilution of Lugol's iodine for tissue squash diagnoses. Since then, different studies have used a variety of iodine dilutions, but most increased the strength of iodine to better visualize parasites in tissues. Because NaOH di- gestion removes most obstructing tissue, iodine strength was re- examined for the whole-oyster protocol. Six aqueous dilutions of Lugol's working solution, ranging from |3;11 to (300; 1) (water to iodine), were compared using pre- zoosporangia extracted from the same oyster after incubation in RFTM. More dilute staining increased the resolution of parasite morphological features and reduced the likelihood of miscounting stained artifacts as positive. However, parasites stained with more dilute solution tended to fade faster. With heavily infected samples that require more time to count, aliquots on filter paper must sometimes be restained. Staining intensity appears to depend on the number of prezoosporangia. the amount of residual oyster tissue, the working stain concentration and the duration of stain- ing. Researchers should test a variety of stain concentrations and examine representative stained parasites for the double wall and signet ring appearance, even though many parasites may not ex- hibit internal characteristics after NaOH digestion. Experiments at EPA/GED found that a |25:1| dilution. Ray"s original recommen- dation, was light enough to determine structural features, yet re- mained dark enough during the counting period to distinguish prezoosporangia from the filter paper background. It should also be noted that working solutions of iodine can precipitate, especially if high concentrations arc being used. Bushek ct al. (1994) noted that false-positives could occur from iodine precipitates and suggested that working iodine solutions should be filtered intermittently and/or periodically examined for precipitate. Refrigerator storage of iodine or stained parasite sam- ples is contraindicated because cold temperature increases precip- itation. Parasite Dilution. Identification and Counting In most cases, sample replication will occur at the counting stage of the protocol so homogeneity of parasite suspensions is critical. Samples should be sonicated for 1-2 min if parasite clumps are present and mixed on a vortex between each replicate aliquot. Since dilutions of the sample will be required (in all but the lightest infections) to achieve countable prezoosporangia den- sities, the dilution process must be accurate and consistent. Dilu- 116 Fisher and Oliver tions should be performed such that larger volumes are transferred in several serial dilutions rather than a smaller volume in a single- step dilution. Tenfold (or less) dilutions and transfer volumes of no less than 100 [jiL are recommended to avoid errors due to lack of homogeneity. Immediately before counting, a 47-mm-diameter. 0.22-|jL-pore- size filter paper (mixed cellulose ester composition) is placed on a borosilicate filter fitted to a vacuum aspirator flask and moistened with distilled water. If triplicate counts are not required, a smaller diameter (25-mni) filter paper may be used. Filters that stain with iodine (i.e.. that contain starch) should not be used. The parasite sample must be mixed vigorously before a 100-|j.L aliquot is placed onto the filter paper and aspirated with a small pump. Replicates may be filtered onto the same filter paper, which is then "mounted" on two microscope slides taped together for exami- nation (no coverslip). Prezoosporangia appear dark blue against the white filter paper background using a light microscope at lOOx magnification. An ocular grid should be used to avoid re-counting microscope fields. If <20 prezoosporangia are found in 100 |jlL. then the entire (1-mL) sample should be counted. This can be accomplished by filtering the entire sample or by concentrating the sample with centrifugation and resuspending in a small volume of stam. If the IOO-|jiL aliquot contains >2(X) prezoosporangia. the sample must be diluted (as noted above) and a new aliquot counted. A minimum l()0-|j.L sample volume is recommended for each serial dilution because of the tendency of prezoosporangia to clump. The use of filter paper for counting parasites has some advan- tages over hemacytometers (Choi et al. 1989). glass slides (Bushek et al. 1994) or tissue culture plate wells (Gauthier and Fisher 1990). Not only do the stained parasites contrast sharply with the white filter paper background, but they are coplanar. This reduces problems associated with parasite clumps dislodging he- macytometer and microscope slide coverslips and problems asso- ciated with locating suspended parasites or parasites attached to tissue culture well walls. It is recommended that at least three counts of lOO-p-L aliquots are made and the mean multiplied by the appropriate dilution factor. Infection intensity can be recorded as prezoosporangia g ^ ' wet weight tissue, based on the starting weight of the entire oyster. Counts should be log,,, transformed for statistical analyses since actual counts are generally highly variable among individual oys- ters and will most likely not fit the assumptions of an analysis of variance model. FINALE The described protocol for whole-oyster diagnosis should pro- vide a more accurate and precise quantitation of P. marinus in- fection intensity than do single-tissue, semi-quantitative protocols. Sensitivity and reproducibility, especially at lower infection inten- sities, remain questionable, but this could be improved by further technical refinement. The generation of cultured lines off. mari- nus (Gauthier and Vasta 1993. Kleinschuster and Swink 1993. La Peyre et al. 1993) could be creatively applied to RFTM technology to verify count precision and sensitivity. Although this protocol is recommended as a standard, different steps within the procedure can be or may need to be modified for specific scientific objec- tives, available equipment, time and personnel, personal prefer- ence, or even oyster size and tissue composition. Researchers will want to adapt incubation periods. NaOH digestion periods and stain concentrations to increase the homogeneity of suspensions and enhance morphological recognition of the parasites. Ration- ales for selected steps were discussed here so that modifications and options can be more easily considered. Whether or not this protocol is modified, researchers will want to generate quality assurance objectives. Some provisions will undoubtedly become routine during the diagnostic process, such as checking iodine solution for precipitation and digested samples for parasite clumps. Others need to be included into the experimental design. At least once in a project, samples should be subdivided prior to RFTM incubation to characterize the variability of the entire diagnostic procedure. Similarly, dilution consistency of stained parasites should be verified periodically. Centrifugation effectiveness after RFTM incubation and NaOH digestion should be confirmed by adding stain to supemates. aspirating them onto filter paper and observing for parasites. Whole-oyster diagnoses may best be used for correlation with oyster physiological measurements or for standardizing and veri- fying other techniques. It is. however, more time intensive and labor intensive than are other techniques. We believe, as did Bushek and co-workers (1994). that it will become the most de- fendable technique for many scientific applications. Determination of disease-free status may be more reliable using nucleic acid recognition techniques, such as those being developed by Marsh and Vasta (1995). ACKNOWLEDGMENTS Technical assistance provided by E. Sutton and R. Webb is greatly appreciated. Participation by Rutgers-The State University and Virginia Institute of Marine Science is also appreciated as are critical comments from Drs. H. Burreson. D. Bushek and S. Ford. This is contribution number 919 from the Gulf Ecology Division. National Health and Environmental Effects Research Laboratory. Gulf Breeze. FL. LITERATURE CITED Anderson. R. S. 1996. Interactions of Perkinsiis marinus with humoral factors and hemocytes of Crassostrea virginica. J. Shellfisli Res- 15: 127-134, Andrews. J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marinus and its effects on the oyster industry. Amer. Fish. Soc. Spec. Publ. 18:47-63. Andrews. J. D. 1996. Hhtory of Perkinsus munnus. a pathogen of oysters in Chesapeake Bay 1950-1984. J. Shellfish Res. 15:13-16. Bushek. D.. S. E. Ford & S. K. Allen. Jr. 1994. Evaluation of methods using Ray's fluid thioglycollate medium for diagnosis of Perkinsus marinus infections in the eastern oyster Crassostrea virginica. Annu. Rev. Fi.sh Dis. 4:201-217. Burreson. E. M. & L. M. Ragone Calvo. 1996. Epizootiology of Perk- insus marinus disease of oysters in Chesapeake Bay with emphasis on data since 1985. J. Shellfish Res. 15:17-34. Choi. K. S.. E. A. Wilson. D. H. Lewis. E. N, Powell & S. M. Ray. 1989. The energetic cost of Perkinsus marinus parasitism in oysters: Quantification of the thioglycollate method. J. Shellfish Res. 8:125- 131 Chu. F.-L. E. 19%. Laboratory investigations of susceptibility, int'ectivity Whole-O'i STER Diagnosis of Perkinsus 117 and transmission of PerkinsKS marinus in oysters. J. Shellfish Res. 15:57-66. Dungan. C. F. & B, S. Roberson. 1993. Binding specificities of mono- and polyclonal antibodies to the protozoan oyster pathogen Perkinsus marinus. Dis. Aqiiat. Org. 15:9-22. Fisher. W. S,, J. D. Gauthier & J. T. Winstead. 1992, Infection intensity of Perkinsus marinus disease in Crassosrrea virginica (Gnielin. 1791) from the Gulf of Mexico maintained under different laboratory condi- tions. J. Shellfish Res. 11:363-369. Fisher. W. S.. L. M. Oliver, E. B. Sutton. C. S. Manning & W. W. Walker. 1995. Exposure of eastern oysters to tributyltin increases the severity of Perkinsus marinus disease. Natl. Shellfish. Assoc. Meet- ing. San Diego (Abstract). J. Shellfish Res. 14:265-266. Ford, S. E. 1996. Range extension by the oyster parasite Perkinsus mari- nus into the northeastern United States: Response to climate change. J. Shellfish Res. 15:45-56. Gauthier, J. D. & W. S. Fisher. 1990, Hemolymph assay for diagnosis of Perkinsus marinus in oysters Crassoslrea virgmica (Gmelin, 1791 ) J. Shellfish Res. 9:367-371. Gauthier, J. D. & G. R. Vasta. 1993. Continuous in virro culture of the eastern oyster parasite Perkinsus marinus. J Inveriehr. Pathol. 62: 321-323. Goggin, C. L. & R. J. G. Lester. 1987. Occurrence oi Perkinsus species (Protozoa, Apicomplexa) in bivalves from the Great Barrier Reef. Dis. Aqual. Org. 3:113-117. Goggin, C. L.. K. B. Sewell & R. J. G. Lester. 1989. Cross-mfcction experiments with .Australian Perkinsus species, Dis. Ai/ual. Org. 7: 55-59. Kleinschuster. S. J. & S. L. Swink. 1993. A simple method for the in rilro culture of Perkinsus marinus. Naulilus 107(2):76-78. La Peyre, J. F.. M. Faisal & E. M. Burreson. 1993. In vitro propagation of the parasite Perkinsus marinus. a pathogen of the eastern oyster, Crassostrea virginiea. J. Euk. Microbiol. 40:304-310. Lester, R. J. G. & G. H.. G. Davis. 1981. A new Perkinsus species (Api- complexa, Perkinsca) from the abalone Haliotis ruber. J Inveriehr. Pathol. 37:181-187. Lewis, D. H., K.-S. Choi & E. N. Powell. 1988. Technique for purifying PfrimiUi man'nui hypnospores. Natl. Shellfish. Assoc. Meeting, New Orieans (Abstract). J. Shellfish Res. 7:204. Mackin, J G. 1951. Histopathology of infection of Crassostrea virginiea (Gmelin) by Dermocystidium marinum Mackin, Owen, and Collier. Bull. Mar. Sci. GulfCaribh. 1:72-87. Mackin, J. G. 1962. Oyster disease caused by Dermocystidium marinum and other microorganisms in Louisiana. Inst. Mar. Sci. Univ. Te.x 7:132-229. Mackin, J. G. & J. L Boswell. 1956. The life cycle and relationships of Dermocystidium marinum. Proc. Natl. Shellfish. Assoc. 46:112-115. Mackin, J. G.. H. M. Owen & A. Collier. 1950. Preliminary note on the occurrence of a new Protistan parasite, Dermocystidium marinum n. sp. in Crassostrea virginiea (Gmelin). Science 111:328-329. Margolis. L., G. W. Esch. J. C. Holmes. A. M. Kuris & G. A. Schad. 1982. The use of ecological terms in parasitology. J. Parasilol. 68: 131-133. Marsh, A. G. & G. R. Vasta. 1995. Parasite-specific PCR amplification of an intergenic mtDNA domain of Perkinsus marinus in oyster he- molymph: A rapid and sensitive assay for infection screening. Natl. Shellfish. Assoc. Meeting, San Diego (Abstract), J. Shellfish Res. 14:271. Oliver, L. M., W. S. Fisher, E. M. Burreson, L. M. Calvo, S. E. Ford & J. Gandy. 1996. Perkinsus marinus tissue distribution and seasonal variation in oysters (Crassostrea virginiea) from Florida, Virginia and New York. Proc. Nail. Shellfish. Assoc. (Abstract) In press. Overstreel. R. M. 1978, Marine maladies? Worms, germs and other sym- bionts from the northern Gulf of Mexico. Mississippi-Alabama Sea Grant Consortium Publication, Blossman Printing, Inc., Ocean Springs. Mississippi. 139 pp. Paynter, K T. 1996. The effects of Perkinsus marinus infection on phys- iological processes of the eastern oyster, Crassostrea virginiea. J. Shellfish Res. 15:119-125 Perkins, F. O. 1988. Structure of protistan parasites found in bivalve mol- luscs. Amer. Fish. Soc. Spec. Publ. 18:93-111. Perkins, F. O. 1993. Infectious diseases of molluscs. In: J. A. Couch & J. W. Foumie (eds.). Pathobiology of Marine and Esluarine Organ- isms. CRC Press. Boca Raton. Florida, pp. 255-287 Perkins. F. O. 1996. The structure of Perkinsus marinus (Mackin. Owen & Collier. 1950) Levine. 1978. with comments on taxonomy and phy- logeny of Perkinsus spp. J. Shellfish Res. 15:67-87. Prokop. J. F. 1950. Infection and culture procedures employed in the study of Dermocxslidium marinum. Project Nine Report to Texas A&M Research Foundation. Ragone Calvo, L. M. & E. M. Burreson. 1994. Characterization of over- wintering infections of Perkinsus marinus (Apicomplexa) in Chesa- peake Bay oysters. J. Shellfish Res. 13:123-130. Ray, S. M. 1952a. .\ culture technique for the diagnosis of infection with Dermocystidium marinum Mackin, Owen, and Collier in oysters. Sci- ence 116:360-361. Ray, S. M. 1952b. A culture technique for the diagnosis of infection with Dermocystidium marinum in oysters. Natl. Shellfish. Assoc. Conven- tion Addresses pp. 9-13. Ray, S. M. 1954a. Studies on the occurrence of Dermocystidium marinum in young oysters. Natl. Shellfish Assoc. Convention Addresses 1953: 80-92. Ray, S. M- 1954b- Biological studies of Dermocystidium marinum. Rice Institute Pamphlet, Houston, 114 pp. Ray, S. M. 1966a. Notes on the occurrence of Dermocystidium marinum on the Gulf of Mexico coast during 1961 and 1962 ( 1963 proceedings). Proc. Natl. Shellfish. Assoc. 54:45-54. Ray, S. M. 1966b. A review of the culture method for detecting Dermo- cystidium marinum. with suggested modifications and precautions (1963 Proceedings). Proc. Nail. Shellfish. Assoc. 54:55-69. Ray, S. M.. J. G. Mackin & J. L. Boswell. 1953. Quantitative measure- ment of effect of disease caused by Dermocystidium marinum. Bull. Mar. Sci. GulfCaribb. 3:6-33. Schmidt. G. D. & L. S. Roberts. 1977. Foundations of Parasitology. C V. Mosby Co.. St. Louis. 604 pp. Soniat. T. M. 1996. Epizootiology of fcr/tmiHimormin disease of eastern oysters in the Gulf of Mexico. J. Shellfish Res. 15:35-43. Soniat. T. M. & M. L. Koenig. 1982. The effects of parasitism by Per- kinsus marinus on the free amino acid composition of Crassostrea virginiea mantle tissue. J. Shellfish Res. 2:25-28. Stein. J. E. & J. G. Mackin. 1957. An evaluation of the culture method used in deteniiining the intensity of Dermocystidium marinum in the oyster C. virginiea. Texas A&M Res. Found. Project 23, Tech. Rept. 22:1-5. Journal of Shellfish Research. Vol. 15. No. I. II*)-!:?. 1^96. THE EFFECTS OF PERKINSUS MARINUS INFECTION ON PHYSIOLOGICAL PROCESSES IN THE EASTERN OYSTER, CRASSOSTREA VIRGINICA KENNEDY T. PAYNTER Departmenl of Zoology University of Maty land College Park. Maryland 20742 ABSTRACT Although Perkinsus inarinus infections have been associated with high mortalities in populations of the eastern oyster Crassostrea virginica for several decades, the pathological mechanism(s) by which death is induced is unclear. Physiological changes in the oyster associated with P. marinus infection are not well studied. Infections typically cause significant reductions in growth rate and several studies have shown reductions in condition index as well. The reduction in condition index may be the result of a perturbation in free amino acid metabolism caused by infection. Further, the effects on free amino acid metabolism may be associated with parasite-induced changes in mitochondrial function. A significant acidosis has also been shown to occur in the hemolymph of infected oysters which may affect general tissue functions. However, changes in oxygen consumption and clearance and assimilation rates of whole oysters have not been correlated with increasing infection. Finally, reproductive capacity may be reduced by P. marinus infection . KEY WORDS: Perkinsus marinns, Crussoslrea viri^iniva. physiology INTRODUCTION Most of the research on the sublethal effects of diseases on bivalves has pertained to growth and mortality. Even though phys- iological data are essential to determine the effects of diseases on oyster populations, very little research regarding the sublethal, physiological effects of diseases has been conducted until recently. However, in the past few years several studies have appeared that provide a beginning to understanding the physiological conse- quences oi Perkinsus marinus parasitism in CrassDstrea viri;inica. Oysters are typically stricken with P. marinus infections in high-salinity (>15 ppt) waters from the Chesapeake Bay south throughout the Gulf Coast states. However, recently. Perkinsus infections have been found in oysters as far north as Massachusetts (Ford 1992) and m Maryland oyster beds at low salinity (< 10 ppt) (Smith and Jordan 1992). Shortly after infection oyster growth usually ceases or is greatly reduced (Andrews 1961. Paynter and Burreson 1991) but large-scale mortality in a population typically does not occur until the second summer of infection. Mortality is associated with high summer water temperatures and is greatest in August or September, depending on the latitude. The typical sce- nario of disease infection, lengthy progression, and eventual death suggests that infection produces significant initial sublethal effects and causes cumulative physiological damage which becomes most acute at high temperatures when metabolic demand in both oyster and parasite is likely highest. The physiological effects of P. marinus infection are most apparent as a reduction in growth rate. However, little is known about the ultimate causes of oyster mortality. Since the protozoan was first described, scientists have observed histological effects which lead to postulation that lytic and cytotoxic agents are pro- duced by the parasite (Mackin 1962, Mackin and Ray 1954), but these results do little to identify the effects on physiological func- tions of the various host tissues and how they may be impaired. Importantly, Mackin (1962) observed that infected oysters could not maintain valve closure as long as uninfected oysters. The inability to maintain valve closure makes oysters much more vul- nerable to predation and limits their ability to tolerate rapid changes in salinity since valve closure is the first defense against osmotic shock. Furthermore, much of the oyster's ability to sur- vive in the estuary is based on cellular adaptation to the environ- ment. How these sophisticated cellular mechanisms are affected has only been recently examined. Oysters such as C. viri>inica are physiologically complex or- ganisms. They dwell in an cstuarine habit that exposes them to a wide range of environmental stresses including high and low tem- peratures, very little or no oxygen for extended periods of time, and rapid salinity changes of 10 ppt or more. To survive these extremes, C. virginica has evolved a sophisticated series of cel- lular and metabolic mechanisms which either neutralize or avoid potentially fatal environmental changes. One of the most well- studied mechanisms is cellular volume regulation in response to a change in salinity (Lynch and Wood 1966, Paynter et al. 1995). When the ambient salinity increases, oyster cells accumulate free amino acids (FAA) to offset the increasing extracellular osmotic pressure. With osmotic pressures nearly equal on both sides of the cell membrane the cell does not lose water and shrink, and can therefore remain functional. Similarly, when oxygen is depleted from ambient water, the oyster consumes less oxygen, lowering its own metabolic rate (Hammen 1980). It makes the chemical energy necessary for survival through alternative metabolic processes which require less oxygen. In this way the cells can continue to function at a low rate in the absence of oxygen. Recent research has focused on the possible effects of P. mari- nus on these sophisticated mechanisms and other physiological characteristics of C. virginica. In the following section I shall review some of the advances made over the last few years includ- ing general effects on growth and condition as well as the effects of infection on FAA concentrations in oysters, differences in mi- tochondria isolated from infected and noninfectcd oysters, the acid-base physiology of oyster hemolymph. and the relationship between P. marinus infection and physiological energetics. These important studies shed light on the mechanisms through which Perkinsus infests and kills oysters and allow us to better under- stand how to prevent or control infections and mortality. 119 120 Pavnter Growth. Condition and Mortality Andrews (1961) used an underwater weighing technique to show that individual oysters that acquired P. maniuis infections exhibited greatly reduced growth. The underwater weighing tech- nique measured shell deposition so that somatic tissue growth, or the effects of infection thereon, could not be assessed. Paynter and Burreson (1991) reported very similar observations on whole pop- ulations of oysters; growth of the whole population was greatly reduced even though less than 100% of the population were diag- nosed as infected. Similar to the study by Andrews, shell height (a measure of shell production rather than somatic tissue) was mea- sured. The difficulties in assessing effects on somatic growth lie in the seasonal variation of tissue weight in oysters; typically dry tissue weight declines in late summer even without infection. Fur- thermore, Hilbish (I9S6) and Borrero and Hilbish (1988) have shown that shell and soft tissue growth in mussels is not always similar and that temporal differences exist between shell and soft tissue growth periods. It is clear, however, that shell deposition is greatly reduced by P marinus infection, subsequently limiting somatic tissue growth. Most studies measuring the effects of P. munims have used a condition index, dry tissue weight per unit shell volume, as a measure of somatic growth or health (Menzel and Hopkins 1933. Craig et al. 1989. Gauthier et al. 1990. Crosby and Roberts 1990. Burreson 1991. Paynter and Burreson 1991. Chu and La Peyre 1993a, Dittman 1993). This measure typically declines after spawning and during hot summer months in most regions and most studies have shown a negative correlation between condition index and P. nuinmis infection (Gauthier et al. 1990. Crosby and Rob- erts 1990. Burreson 1991. Paynter and Burreson 1991. Dittman 1993. Volety and Chu 1994). However, other studies (Chu and La Peyre 1993b. Chu et al. 1993. Newell et al. 1994) have revealed no relationship between P. marinus infection and condition index. Laboratory exposures of relatively short duration may not allow enough time for condition to become reduced, but they neverthe- less reveal that reduction in condition is not always directly asso- ciated with increasing infection intensities. Free Amino Acid Metabolism C. virginica is a euryhaline species capable of acclimating to wide changes in ambient salinity. Cell volume is controlled by regulating a large, intracellular FAA pool and the quaternary am- monium compound glycine belaine to offset changes in extracel- lular osmotic pressure, i.e.. this oyster is an osmoconformer (Pierce et al. 1992). The time course of amino acid accumulation has been measured in oysters exposed to increased salinity (Paynter et al. 1993) and appears to be typical of other bivalves. In ribbed mussels, for instance, alanine rapidly accumulates and reaches high levels immediately after a hyperosmotic stress. As acclimation proceeds, the glycine concentration rises and within a few days replaces alanine as the major osmotic effector. During the next several days to weeks at high salinity, proline typically appears as a transient peak, beginning to rise slowly alter the alanine accumulation peaks and declining as taurine accumulates. Taurine usually becomes the major osmotic effector, often com- prising as much as 70% of the FAA pool (Baginski and Pierce 1977). The ability to regulate intracellular amino acids in this way allows C. virginica to inhabit estuaries such as Chesapeake Bay. While the physiological effects of protozoan parasitism have been addressed by a few studies (Newell 19X3. Barber et al. 1988a, 1988b, Ford and Figueras 1988, Newell and Barber 1988), none have examined salinity tolerance. Heavily infected oysters appear wasted, watery, and translucent, and the ratio of whole wet tissue weight to dry tissue weight is increased (Paynter and Bur- reson 1991). In addition, oysters from various locations within Chesapeake Bay had smaller intracellular free amino acid pools, almost no glycine betaine. and reduced salinity tolerances com- pared with conspecific oysters from several locations along the Atlantic Coast trom Georgia to Cape Cod (Pierce et al, 1992). It is likely that all of the Chesapeake Bay oysters in that study were parasitized with P. marinus. These observations, together with the observed reductions of morbidity and mortality amongst parasit- ized oysters in lower salinities (Andrews 1988, Burreson and An- drews 1988. Paynter and Burreson 1991 ). suggest that P manmis infection may have an impact on the salinity tolerance mechanisms of the oyster, a hypothesis suggested many years ago by Soniat and Kocnig (1982). Paynter et al. (1993) examined the effects of both P. marinus parasitism and environmental salinity on intracellular FAA con- centrations of oyster tissues during a cycle of infection in the held. 300 20 40 60 80 100 120 Day Figure 1. Concentration of taurine, glycine, and total VW in gill tissues from oysters held at high (20 ppt, open square)- and low (8-12 ppt. closed circle)-salinity sites. Transfer of oysters from low to high salinity occurred on day 0. \n increase from 8 to 12 ppt occurred at the low-salinity site between days 10 and 20. .\sterisks denote infection of oyster groups by F. marinus at high salinity. From Paynter et al. 1995. Physiological Effects of Perkinsvs on Oysters 121 I In that study, the FAA levels in gill tissues changed after transfer to the high-salinity site (Fig. 1 ). essentially as predicted by earlier laboratory studies on other bivalve species. Overall, the total FAA increased to a peak around 500 jxmol/g dry wt within .^ d and remained constant until the day 85 sample when protozoan infec- tions were first detected and total FAA levels declined. Taurine concentrations at days 85 and 105 were significantly lower than the levels exhibited before infection. Remarkably, glycine and total FAA declined to levels that were not different from those of the low-salinity oysters. The amino acid levels in tissues from the oysters held at low salmity stayed constant after the period be- tween days 10 and 20 when the total FAA level went up. presum- ably in response to a 4 ppt salmity increase at the site. The mean levels of glycine, taurine, and total FAA remained constant at the low-salinity site for the remamder of the study period. The levels of FAA in the oysters transferred to high salmity in this study were significantly different among three phases (unin- fected, lightly infected, and heavily infected) following salinity acclimation. Mean levels of the major amino acids of the FAA pool — taurine, glutamate. and glycine — and the minor compo- nents— threonine, serine, and (i-alanine — were all lower in the infected groups compared to the uninfected groups. All of the FAA, except taurine and giutamine. were significantly lower even with light infection intensity, producing a 247i decline in the total FAA pool. Glycine and p-aianinc levels declined further as infec- tion intensity increased and taurine levels were substantially re- duced in the heavily infected oysters. Both giutamine and alanine levels increased in the heavily infected group. Overall, the total FAA pool in the oysters at the high-salinity field site declined by 40'7f between the acclimated uninfected phase and the heavily infected phase. In summary, intracellular FAA levels were much lower in the gills of the groups of high salinity-adapted oysters infected by P. mannus. As the infection intensified in the group, the amino acid pool declined by 409f of its original level largely due to taurine decreases in all of the oysters, including those scored as uninfected by our fluid thioglycollate-based diagnostic method. Since taurine levels in oysters or mussels acclimated to a particular salinity remain at the acclimated level unless a subsequent salinity change occurs (Soniat and Koenig 1^82, Baginski and Pierce 1977. Bishop et al. 1983, Pierce et al. 1992) and given the lack of a similar decline in taurine in the low-salinity oysters, the reductions in taurine and other FAA in the oysters are likely related to pro- tozoan infection rather than some type of seasonal change. Since oysters are osmoconformers. a reduction of 33* in intracellular FAA must be compensated for by the elevation of other intracel- lular solutes. At present, these solutes are not known but inorganic ions such as Na^, K*, and CP are obvious possibilities. An increase in intracellular ion concentration of this magnitude is likely to produce negative physiological effects (Yancey et al. 1982). The decreased FAA concentration might be caused by per- turbations either in the synthesis of FAA important for salinity tolerance or in the membrane characteristics which keep the FAA inside the cell once they are synthesized. The intracellular ammo acids that are utilized for salinity tolerance are synthesized largely by the mitochondria (Paynter et al. 1984, Pierce et al. 1992). Once synthesized, the amino acids are transported to the cytosol where their intracellular concentration is regulated by the permeability control mechanisms of the cell membrane (Pierce and Politis 1992). Therefore, the presence of P. maninis might affect the permeability control mechanisms of the cell membrane or syn- thetic mechanisms in the mitochondria. In addition to the effects on the osmolytcs. impaired mitochondrial function could result in a reduction in ATP production, which could account for the re- duction of growth observed in the presence of the parasite. Mitochondrial Metabolism Pierce et al. ( 1992) found that oysters from a variety of locales within Chesapeake Bay had salinity tolerances that were more narrow than the tolerances of oyster populations elsewhere along the Atlantic Coast. The basis of this difference was that the FAA pool of Bay oysters was smaller and composed of different amino acids than that of the Atlantic oysters, agreeing with the results reported by Paynter et al. ( 1995) summarized above. In addition, the Atlantic oysters had substantial intracellular concentrations of glycine betaine not present in the cells of Bay oysters. These differences in osmolyte composition between Chesapeake and At- lantic oysters no doubt account for the salinity tolerance differ- ences. Furthermore, since both the amino acids used in cellular osmoregulation and glycine betaine are synthesized in the mito- chondria (Paynter et al. 1984). the differences between Chesa- peake and Atlantic oysters may reside in the mitochondria. In addition, since all of the Bay oysters studied have been parasitized with P. marinus. it is possible that the differences are due to the parasite rather than to genetics. The respiratory control ratios (RCRs) of mitochondria from Bay oysters are often higher than those of mitochondria from Atlantic oysters (Fig. 2). In addition, the Bay oyster mitochondria give highest RCRs with malate as a substrate while the Atlantic mitochondria prefer a-ketoglutaric acid. The basis of this coupling ratio difference is not clear at present, but at least suggests that the energy metabolism of Bay and Atlantic animals is different and that the difference may lie with the control of the kinetics of various steps in the Kreb's cycle. In addition. Pierce et al. ( 1995) have shown that the uptake of choline, a precursor of the osmotically active compound glycine betaine. by mitochondria isolated from Chesapeake Bay oysters is significantly lower than that by mitochondria from Atlantic con- specifics. While infection levels were not part of this study, all of the Chesapeake oysters tested from the experimental groups were RCRs trom Atlantic and Bay oyster gill mltoctiondria adapted to 350 mosm 5 o e I 0 J ■ Atlantic n Chesapeake glu a-kg Substrates Figure 2. Respiratory coupling ratios of mitochondria isolated from Atlantic and Chesapeake Ba> oysters acclimated to low salinity. His- togram bars are the means of at least 10 measurements. Error bars indicate standard errors. From Pierce et al. 1992. 122 Paynter infected with P . maruuis while most of the Atlantic oysters were not. This suggests that mitochondrial metabolism may be affected by P. mariiius infection. The initial accumulation of FAA in response to hyperosmotic shock is the result of a complex biochemical regulatory process that allows oyster cells to route carbon and nitrogen in a very specific way (Bishop et al. 1983). Since the biochemical pathways used to respond to hyperosmotic stress may be the same as. or very similar to, the pathways involved in hypoxic tolerance (Baginski and Pierce 1975). it is possible that the ability of oysters to tolerate hypoxia, which is common in Chesapeake Bay (MacKiernan 1987), might also be diminished by P. marinus infection A re- duced tolerance of hypoxia could lead to the large-scale mortalities that occur in Chesapeake Bay oysters during exposure to hypoxia, which occurs frequently during the summer months over many oyster bars. Hypoxia Tolerance and Acid-Base Balance Hypoxia or anoxia causes not only the obvious stress associated with the lack of oxygen but also causes a general decrease in tissue pH caused by an increase in CO-,. These changes can have harmful effects on the general well-being of organisms and affect many aspects of normal physiological performance. When oysters are air exposed, they close their valves and the oxygen contained in the water trapped between the valves is quickly exhausted. The oyster tissues then become hypoxic and hypcrcapnic. Infected oysters cannot hold their valves closed for as long as uninfected oysters when aerobically exposed (Fig. 3). This is thought to be due to the stress induced by infection, but the exact nature of the stress has never been addressed. Dwyer and Burnett ( 1996) have studied the acid-base physiology of oysters and the effects of P . marinus on acid-base balance and discovered important correlations between infection and acidosis in oysters. Dwyer and Burnett ( 1996) showed that the normal pH of oyster hemolymph is around 7.7 while oysters infected with P. inannus show significant acidosis (pH 7.2; Fig. 4). Furthermore, they showed that minimal hypoxic stress caused a large decline in he- molymph pH in both infected and noninfected oysters but that Pco, 40 Pco, 15 20 0 2 4 6 8 10 Time to Gape (ciays) Figure 3. Relationship between weighted prevalence of P. marinus infection and time to gape of oysters held undisturbed out of water at room temperature. Regression analysis revealed a strong negative cor- relation between weighted prevalence and time to gape (P - 0.001). Crassjstrea virginica 2fC Pco, 5 Pco, 2 pH Figure 4. \ pH-HCO, diagram showing the acid-base status of oys- ter hemol\mph at 0, 5, and 24 hr of air exposure at 2rC. PCO, isopleths (curved lines) are given in torr. In vitro bulTer lines are shown as dashed lines. Circles represent uninfected oysters: triangles represent oysters with high infections. Values are means ± SE. N for each experiment ranged from 22 to 56. From Dwyer and Burnett 1996. infected oysters incurred a steeper decline. The pH of hemolymph of infected oyster dropped to 6.7 after 5 hr of hypoxic stress compared to a decline to 7.3 in healthy oysters. This response could result in large differences in nutrient absorption or retention, blood cell function (see Anderson 1996). oxygen consumption, and metabolic efficiency between infected and uninfected oysters. Indeed, it could be associated with the loss of FAA from cells or an increased rate of parasite growth. It is likely that this acidosis is associated with the inability of infected oysters to remain closed as long as uninfected oysters. In fact. Dwyer and Burnett (1996) have shown that adductor muscles of infected oysters contain sig- nificantly lower amounts of glycogen than do uninfected oysters and that heavily infected oysters have less glycogen than lightly infected individuals. This suggests a direct correlation between disease and an oyster's ability to perform ecologically critical tasks such as keeping its valves closed (see Fig. 3). These kinds of studies bring us to a closer understanding of the nature of mortal injury inflicted by the parasite. Physiological Energetics Newell ( 1985) reported that the feeding rales of eastern oysters were significantly reduced when they were infected by the parasite Haptosporidiiim nelsoni (MSX). Such reduced feeding activity resulted in lower amounts of glycogen being sequestered (Barber et al. 1988a). resulting in a reduction in condition index. Energy allocation by MSX-infested oysters to gametogenesis was also disrupted, resulting in significantly inhibited gametogenesis dur- ing the spring (Ford et al. 1990). The effects of P. marinus infec- tions on the physiological functions of feeding, metabolic energy expenditure, and assimilation efficiency in eastern oysters have only recently been studied. Using a combination of field and laboratory experiments to study the effects of Perkinsus infections on C. virginica. Newell et al. (1994) have shown that P. marinus infection has a surprisingly small effect on most aspects of feeding physiology and metabo- lism. In that study, oxygen consumption, clearance rates, condi- Physiological Effects of Perkinsus on Oysters 123 tion index, and assimilation rates were measured over a 2-year cycle of growth and infection at three sites in Chesapeake Bay. When the oysters were transferred from low to higher salinity sites at the initiation of the experiments, feeding rates increased sharply and reniamed high for 1 to 2 months but declined rapidly in con- cert with the first detection of P. marinus infections in the exper- miental groups. Surprisingly, however, further declines in the clearance rates of the oysters did not occur in association with progression of disease. In contrast, Mackin and Ray (1954) showed that fecal and psuedofecal production in moderate and heavily infected oysters was less than half that of uninfected oys- ters. Other aspects of the physiological energetics were unaffected by P. mwiiuis infection. Oxygen consumption did not change between uninfected and infected oysters, even in oysters heavily infected and within weeks of succumbing to the disease. Condition index has been reported to decline with infection (Craig et al. 1984. Gauthicr et al. 1990. Crosby and Roberts 1940, Paynter and Burreson 1491, Dittman 1993) but was not associated with infec- tion level in Newell et al. ( 1444). .Most surprisingly, assimilation rates remained unchanged by P . marinus infection even though the primary portal of entry by this parastie is thought to be through the intestine. Clearly, the nature of physiological damage caused by P mariiiiis infections cannot yet be fully understood. Reproductive Capacity Parasitism by H . nelsoiii has been shown to significantly affect reproduction in oysters. Gametogenesis is apparently inhibited by infection-induced disruptions in carbohydrate metabolism (Barber et al. 141S8b. Ford et al. 1490) which lead to a reduction in fecun- dity (Barber et al. 1988a). Given these observations one might expect that P. marinus infections would also have significant del- eterious effects on reproduction in the oyster. However, the effects of P. marinus infection on gamete production seem to be much less direct. Although fecundity or reproductive condition has not been di- rectly studied in relation to P. marinus infection as it has with H. nelsoni (Barber et al. 1988a), Cox and Mann ( 1992) have shown that reductions in reproductive activity in oyster populations in the James River, VA, have coincided with increases in P. marinus prevalence over the last few years. It also seems likely that Per- kinsus infection may cause perturbations in gonad development since Ragone Calvo and Burreson ( 1994) showed that P. marinus parasites survive overwintering and can develop into substantial infections soon after temperatures increase, which may coincide with gonadal maturation and spawning. Kennedy et al. (1995) showed that P. marinus infections ac- quired during the previous year did not have a deleterious effect on reproduction the following year. Similarly. Dittman (1993) showed that oysters with light first-year infections had percent gonad areas that were similar to those of uninfected oysters. How- ever, percent gonad areas in oysters with heavy infections were significantly lower, indicating a significant negative impact of Perkinsus infection on reproductive capacity. Ray et al. (1953) and Kennedy et al. (1995) reported significant reductions in re- productive output, in terms of numbers of eggs produced, of oys- ters heavily infected with P. marinus. In contrast, eggs from P. mariniis-infected individuals were not smaller than eggs from un- infected animals and the lipid content of eggs from infected oysters was no different from that of eggs of uninfected oysters (Kennedy et al. 1995). Thus, it appears that heavy P. marinus infection may have some deleterious effect on reproduction but that perhaps the oyster can shunt energy from growth (which is reduced even with light infections) to gametogenesis to minimize the effects of in- fection on egg quality. SUMMARY Infection of the eastern oyster by P. marinus induces a number of significant changes in the physiology of the oyster. Given the changes in the hemolyniph pH associated with infection, one would expect nearly all cell-mediated functions, including ciliary beating, respiration, absorption of nutrients, and excretion of waste products, to be altered. Certainly acidosis could inhibit cal- cification and shell deposition, accounting for the cessation of shell growth associated with infection. It could also alter mem- brane characteristics to the point where amino acid uptake was retarded and it may account for changes in blood cell function as described by Anderson (1996). However, given the expected re- sponse of general acidosis, the observations of Newell et al. (1994), which show little or no effect on oxygen consumption, clearance rate (a measure of ciliary action of the gills), or food assimilation, are startling. These contradictory observations only serve to demonstrate the need for a better understanding of the biochemical, pharmacological, and physiological effects of P. marinus infections on oysters. La Peyre and Faisal ( 1996) have shown that P. marinus cells in culture produce extracellular proteases. These proteases are thought to play a role in damaging host tissue, protecting the parasite from host immune response, and perhaps enhancing the parasites" ability to replicate within the host. General changes in the host physiology such as a decline in hemolymph pH may enhance the cytotoxic activity of such parasite-produced chemical agents. It is important to note that the physiological effects off. mari- nus infection on oysters may differ between physiological races of C. viri;inica. Several studies (Bushek and Allen 1996. Paynter and Burreson 1991. Pierce et al. 1992, Brown et al. 1994) have shown that different intraspecific populations of oysters respond differ- ently to P. marinus infections. Therefore, a level of infection that would induce mortality in one population may not induce mortality in another population (Brown et al. 1994). This makes it more important to understand the mechanisms of pathology that result in mortal itv in oysters. LITERATURE CITED Anderson, R. S. 19%. Interactions of Perkinsus marinus with humoral fac- tors and hemocytes of Crassoslrea virginica. J. Shellfish Res. 15:127-134. Andrews. J D. 1961. Measurement of shell growth in oysters by weighing in water. Proc. Natl. Shellfish Assoc. 52:1-11. Andrews, J. D. 1988. Epizootiology of the disease caused by the oyster pathogen, Perkinsus marinus and its effects on the oyster industry. Amer. Fish. Soc. Spec. Publ. 18:47-63. Baginski. R. M. & S. K. Pierce. 1975. Anaerobiosis: a possible source of osmotic solute for high salinity acclimation in marine molluscs. J. Exp. Biol. 62:589-598. 124 Paynter Baginski, R. M. & S. K. Pierce. 1977. The time course of intracellular free amino acid accumulation in tissues of Modiolus demissus during high salinity adaptation. Comp. Biochem. Physiol. 57(4A):407^13. Barber, B. J., S. E. Ford & H. H. Haskin. 1988a. Effects of the parasite MSX {Haplosporidiiim nelsoiii) on oyster iCnissoslrea virginica) en- ergy metabolism. I. Condition index and relative fecundity. J. Shellfish Res. 7:25-31. Barber. B. J., S. E. Ford & H. H. Haskin. 1988b. Effects of the parasite MSX {Haplosporidium nelsoni) on oyster (Crassosirea virginica) en- ergy metabolism. II. Tissue biochemical composition. Comp. Bio- chem. Physiol. 91A:603-608. Bishop. S. H.. L. L. Ellis & J. M. Burcham. 1983. Amino acid metab- olism in molluscs. In: K. M. Wilbur (Ser. Ed. I & P. W. Hochachka {Vol. Ed.). The Mollusca (2nd edition), vol. 1. Academic Press. Inc.. New York. pp. 243-327. Borrero, F. J. & T. J. Hilbish. 1988. Temporal vanation in shell and soft tissue growth of the mussel Geukensni demissu. Mar. Ecol. Prog. Ser. 42:9-15. Brown. B. L.. A. J. Butt & K. T, Paynter. 1994. Performance vanation among native and selectively-bred eastern oyster strains in North Caro- lina. J. Shellfish Res. 13:292. Burreson. E. M. 1991. Effects of Perkinsiis marimis infection in the east- em oyster, Crassosirea virginica: I. Susceptibility of native and MSX- resistant stocks. J. Shellfish Res. 10:417-423. Burreson, E. M. & J. D. Andrews. 1988. Unusual intensification of Ches- apeake Bay oyster diseases during recent drought conditions. Oceans 88 Proc. Vol. 3:799-802. IEEE Cat. No. 88-CH2585-8. Bushek. D. & S, K. Allen. Jr. 1996. Races of Perkinsiis marinns. J. Shellfish Res. 15:103-107. Chu. F. E. & J. F. La Peyre. 1993a, Development of disease caused by the parasite. Perkinsiis mariniis and defense-related hemolymph factors in three populations of oysters from the Chesapeake Bay. USA. J. Shellfish Res. 12:21-27. Chu, F. E. & J. F. La Peyre. 1993b. Perkinsiis marinus susceptibility and defense-related activities in eastern oysters, Crassosirea virginica: temperature effects. Dis. Aqiiat. Org. 16:223-234. Chu, F. E.. J. F. La Peyre & C. S. Burreson. 1993. Perkinsiis marinus infection and potential defense-related activities in eastern oysters. Crassosirea virginica: salinity effects. J Inveriehr. Palhol. 62:226- 232. Cox. C. & R. Mann. 1992. Temporal and spatial changes in fecundity of eastern oysters. Crassosirea virginica (Gmelin. 1791). in the James River, Virginia. J. Shellfish Res. 11:49-54. Craig. A.. E. N. Powell. R. R. Fay & J. M. Brooks. 1989. Distribution of Perkinsus marinus in Gulf coast oyster populations. Esluaries 12:82- 91. Crosby. M. P. & C. F. Roberts. 1990. Seasonal infection intensity cycle of the parasite Perkinsus marinus (and an absence of Haplosporidium spp.) in oysters from a South Carolina salt marsh. Dis. Ac/uai. Org. 9:149-155. Dittman. D. E. 1993. The quantitative effects of Perkinsus marinus on reproduction and condition in the eastern oyster, Crassosirea virginica. J. Shellfish Res. 12:127. Dwyer, J. J. & L. E. Burnett. 1996. Acid-base status of the oyster, Cras- sosirea virginica. in response to air exposure and to infections by Perkinsus marinus. Biol. Bull. 190:139-147. Ford, S. E. 1992. Avoiding the transmission of disease in commercial culture of molluscs, with special reference to Perkinsus marinus (Dermo) and Haplosporidium nelsoni (MSX). J. Shellfish Res. II: 539-546. Ford S. E. & A. J. Figueras. 1988. Effects of sublethal infection by the parasite Haplosporidium nelsoni (MSX) on gametogenesis. spawning. and sex ratios of oysters in Delaware Bay. USA. Dis. Aqiial. Org. 4:121-133, Ford. S. E.. A. J. Figueras & H. H. Haskin. 1990. Influence of selective breeding, geographic origin, and disease on gametogenesis and sex ratios of oysters. Crassosirea virginica. exposed to the parasite Hap- losporidium nelsoni (MSX). Aqiiaculliire 88:285-301. Gauthier. J. D., T. M. Soniat & J. S. Rogers. 1990. A parasitological survey of oysters along salinity gradients in coastal Louisiana. J. World Aquacidl. Soc. 21:105-115. Hammen. C. S. 1980. Marine Inverlebraies: Comparative Phvsiologv. University Press of New England. Hanover. New Hampshire. 127 pp. Hilbish. T. J. 1986. Growth trajectories of shell and soft tissues in bi- valves: seasonal variation in Mxlilus ediilis L. J. Exp. Biol. Ecol. 96:103-113. Kennedy. V. S.. R. I. E. Newell, G. E. Krantz & S. Otto. 1995. Repro- ductive capacity of the eastern oyster, Crassosirea virginica, infected with the parasite Perkinsus marimis. Dis. Aqual. Org. 23:135-144. La Peyre. J. F. 1996. Propagation and in vilro studies of Perkinsus mari- nus. J. Shellfish Res. 15:89-101. Lynch. M. P. & L. Wood. 1966. Effects of environmental salinity on free amino acids of Crassosirea virginica Gmelin. Comp. Biochem. Phys- iol. 19:783-790. MacKieman. G. B. 1987. Dissolved O.xygen in ilie Chesapeake Bar: Pro- cesses and Effects. Maryland Sea Grant College. College Park, Mary- land. 177 pp. Mackin. J, G. 1962, Oyster diseases caused by Dermocvsiidium mannum and other microorganisms in Louisiana Puhl. Insl. Mar. Sci. Univ. Tex 7:132-229. Mackin. J. G. & S. M. Ray. 1954. Studies on the effect of infection by Dermocystidium marinum on ciliary action in oysters {Crassosirea vir- ginica). Proc. Nail. Shellfish Assoc. 45:168-181. Menzel. R. W. & S. H. Hopkins. 1955. The growth of oysters parasitized by the fungus Dermocystidium marinum and by the trematode Buceph- alus cuculus. J. Parasilol. 41:333-342. Newell, R. I. E. 1985. Sublethal physiological effects of the parasite MSX (Haplosporidium nelsoni) on the oyster Crassosirea virginica. J. Shell- fi.shRes. 5:91-95. Newell, R I. E & B, J, Barber, 1988, A physiological approach to the study of bivalve molluscan diseases, Amer. Eish. Soc. Spec. Puhl. 18:269-285. Newell, R. I. E., K. T. Paynter & E. M. Burreson. 1994. Physiological effects of protozoan parasitism on the eastern oyster Crassosirea vir- ginica: feeding and metabolism. J. Shellfish Res. 13:294. Paynter. K. T.. Jr. & E. M. Burreson. 1991. Effec\& of Perkinsus marinus infection in the eastern oyster Crassosirea virginica II. Disease devel- opment and impact on growth rate at different salinities. J. Shellfish Res. 10:425-431. Paynter. K. T.. L. L. Ellis & S. H. Bishop. 1984, Cellular location and partial characterization of alanine aminotransferase from nbbed mussel gill tissue. J. Exp. Zool. 232:51-58. Paynter. K. T.. S. K. Pierce & E. M. Burreson. 1995. Free amino acid levels in eastern oysters. Crassosirea virginica. infected with Perkin- sus marinus. Mar. Biol. 122:67-72. Pierce. S. K. & A. D. Politis. 1992. Ca"* -activated volume recovery mechanisms. Annu. Rev. Physiol. 52:27^2. Pierce. S. K.. L. M. Rowland-Faux & S. M. O'Brien. 1992. Different salinity tolerance mechanisms in Atlantic and Chesapeake Bay con- specific oysters: glycine betaine and amino acid pool variations. Mar. Biol. 113:107-113. Pierce, S. K.. L. M. Rowland-Faux & B. Crombie. 1995, The mechanism of glycine betaine regulation in response to hyperosmotic stress in oyster mitochondria: a comparative study of Atlantic and Chesapeake Bay oysters. J. Exp. Zool. 271:161-170. Ragone Calvo. L. M. & E. M. Burreson. 1994. Charactenzation of over- Physiological Effects of Perkinsus on Oysters 125 wintering infections of Perkinsus marinus (Apicomplexa) in Chesa- kinsus marinus on the free amino acid composition of Crassoslrea peake Bay oysters. J. Shellfish Res. 13:123-130. virgiiiica mantle tissue. J. Shellfish Res. 2:25-28. Ray. S. M.. J. G. Mackin & J. L. Boswell. 1953. Quantitative measure- ,, , _ . ,, p r- r- /-u inn. ,- c- r ■ j ■. ■' J Volety, A. K. & r. E. Cnu. 1994. Companson of intectivity and patno- nient of the effect on oysters of disease caused by Dermocvstidiiim ■ ■. c . ,. u •. , j ^ <. ' genicity or meront (trophozoite) and prezoosporangia stages of (he marinum. Bull. Mar. Sci. Unit Caribh. 3:6-33. . u n ; ^ oyster pathogen Perkinsus marinus in eastern oysters. Crassoslrea vir- Smith. G. F. & S. J. Jordan. \992. Monitorini> Mar\lantl s Chesaneake ,_ ,. ,^n.. > c, m- t n .ic-,. c^-, gimca (Gmelin, n9\). J. Shellfish Res. 13:521-527. Bay oysters — A Comprehensive Characterization of Moaified Fall Sur- vey Results 1990-1991. Maryland Dept. Nat. Resources CBRM-OX- Yancey. P. H.. M. E. Clack. S. C. Hand, R, D. Bowlus & G. N. Som- 93-3. ero. 1982. Living with salt stress: evolution of osmolyte systems. Soniat. T. M. & M. L. Koenig. 1982. The effects of parasitism by Per- Science 217:1214-1222. J, ninuil inui.i. hemocyle. defense mechanisms, humoral defense mol- ecules, reactive oxygen species, lysozyine. proteases, acid phosphatases, iron-binding proteins, stress proteins, agglutinins BRIEF OVERVIEW OF THE DEFENSE MECHANISMS OF CR.ASSOSTREA MRGIMCA In order to focus on the subject of this chapter, the putative involvement of Crassostrea virginica defense mechanisms in the host response to Perkinsiis marimis Infection, no attempt will be made to present a complete review of the literature on immunity in other bivalve species,- Instead, the protective mechanisms avail- able to the eastern oyster will be mentioned brietly. while most of this contribution will describe what is currently known about the effects that these physiological mechanisms have on P. marinus. and vice versa. Although this parasite has had devastating effects on C. virginica on the East Coast of the United States, surprisingly little is known about specific interactions between it and the de- fense reactions of its host. Throughout this chapter the terms "im- mune mechanisms" and "defense mechanisms" will be used in- terchangeably. Specific or adaptive immunity characterized by the presence of lymphocytes and immunoglobulins cannot be found in oysters; however, cellular and humoral components analogous to those typical of the nonadaptive (nonspecific) immune systems of higher animals are present. Phagocytes, antimicrobial molecules, agglutinins, lysins. and other humoral components comprise the immune systems of invertebrates. Humoral Components A number of lysosomal hydrolases not only participate in the intrahemocytic destruction of microorganisms but also find their way into the cell-free hemolymph of the oyster, Degranulation of hemocytes probably is a major contributing mechanism for the extracellular presence of these enzymes (Foley and Cheng 1977. Cheng et al. 1975. Cheng 1992). These factors are thought to function in the control of infection, and lysozyme is probably one of the most studied examples (McDade and Tripp 1967a. Cheng and Rodrick 1975, Hardy et al. 1977, Steinert and Pickwell 1984). Lysozyme levels show considerable natural variation among oys- ters, as well as variation due to seasonal effects, the presence of parasites, and exposure to xenobioties (Feng and Canzonier 1970, Cheng and Rodrick 1974, Pickwell and Steinert 1984, Chu and La Peyre 1989). Other lysosomal enzymes including acid phos- phatase, aminopeptidase. beta-glucuronidase, and lipase have also been detected in hemolymph (Cheng 1976b, 1978, Cheng and Rodrick 1975, Yoshino and Cheng 1976, Fries 1984, Pipe 1990). The roles of these enzymes in antimicrobial responses are less studied than lysozyme; however, they may degrade surface integ- rity of microorganisms contributing to their recognition and/or destruction by the host. Another widely studied class of oyster humoral factors is the agglutinins (lectins! directed against various saccharide moieties on cell surfaces. These molecules react avidly with mammalian ervthrocytes and have been shown to facilitate their subsequent phagocytosis (Tripp 1966, McDade and Tripp 1967b), It is pos- tulated that their normal in vivo functions may involve extracel- lular recognition and opsonization of bacteria and protozoans (Renwrantz 1983. Fisher and Dinuzzo 1991). and they may also serve as non-self-receptors associated with the hemocyte surface (Vasta et al. 1982. Renwrantz and Stahmer 1983, Vasta et al. 1984. Vasta 1991). There is evidence that naturally occurring oyster agglutinins, as well as plant lectins. (Mullainadhan and Renwrantz 19861. can form bridging molecules between hemo- cytes and non-self-material. However, the lectin-mediated recog- nition of foreignness is variable, dependent on the determinants available on a given foreign particle or cell. The requirement for opsonization in phagocytosis by molluscan hemocytes is not ab- solute; recognition and uptake can proceed in the absence of serum factors, or in hemolymph with no apparent ability to agglutinate the foreign material (Renwrantz and Stahmer 1983). Cellular Components The hemolymph of the oyster contains several hemocyte types that can be identified based on morphology (Cheng 1981; Fisher 1986). The main phagocytic hemocytes are often referred to as granulocytes and hyalinocytes. This distinction is based on relative degrees of granularity and is rather subjective but will have to suffice until functional subsets with unequivocal surface markers 127 128 Anderson are identified. The granular hemocytes are more avidly phagocytic in oysters. The numbers of hemocytes in the circulation of C. virginica increase during P. inanuiis infection (Anderson et al. 1992c); similar increases have been noted in other bivalves stressed by toxicants or infections (Cheng 1988. Renwrantz 1990. Oubella et al. 1993. Coles et al. 1994). The hemocytes are thought to be of central importance in resisting and controlling infections in bivalves by virtue of their ability to ingest and destroy micro- organisms. In addition to the presence of cidal and digestive hy- drolases, oyster ceils can generate antimicrobial reactive oxygen intermediates (ROIs) (see reviews by Adema et al. 1991, Ander- son 1994b). It is likely that C. virginica hemocytes contain my- eloperoxidase (MPO) which is involved in the production of the potent antimicrobial factor hypochlorous acid. MPO has been demonstrated in mussel hemocytes (Schlenk et al. 1991). Zymo- san-stimulated, luminol-augmenled chcmilumincsccncc (CL) is strong in oyster hemocytes. mdicatmg the presence of the MPO/ HiOi/halide antimicrobial pathway. Phenoloxidase may also be present in oyster hemocytes, as it is in the cells of Myiiliis cthilis (Coles and Pipe 1994). Involvement of this enzyme in microbi- cidal, antiparasitic, recognition, opsonization, and cellular com- munication aspects of arthroptcran immunity has long been rec- ognized (Soderhall 1982, 1992). The hemocyte-niediated immune mechanisms of C. virginica in relation to P niariniis infection will be detailed in subsequent sections. AMBIENT ENVIRONMENTAL EFFECTS ON IMMUNE PARAMETERS AND P. MARINVS INFECTION Obviously, the ambient temperature might be expected to in- fluence hemocyte functions and defense-related activities of oys- ters and other ectotherniic organisms. Earlier literature indicated that elevated temperatures were associated with increased pinocy- tosis, migratory activity, and phagocytosis (Feng and Feng 1974, Feng 1965b, Foley and Cheng 1975). Fisher and Taniplin ( 1988) reported that hemocyte rate of locomotion and foreign particle- binding capacity showed positive correlation with temperature. Phagocytosis rates in oysters acclimated to various temperatures (Chu and La Peyre 1993b) also rose as temperatures approached 25°C. These authors found that both the total hemocyte count and the percentage of granulocytes in circulation increased with tem- perature. This could result from concomitant higher heart contrac- tion rate (Feng 1965a) or some other means of mobilization from noncirculating hemocyte pool(s). It is hypothetically possible that increased cell counts result from hematopoiesis, but histological areas of blood cell proliferation have not as yet been described in oysters. Oysters collected in Virginia, and more northern sites, had serum lysozyme levels that were generally higher during win- ter months than in the summer (Feng and Canzonier 1970, Chu and La Peyre 1989): however, Fisher et al. ( 1993) reported higher lysozyme levels in summer months in Gulf Coast oysters, sug- gesting possible regional differences in temperature influences. Lysozyme levels in laboratory-held oysters were also lower at elevated temperatures, but hemagglutinin titers were unchanged (Chu and La Peyre, 1993b). The relationship between environmental temperature and P. marinus incidence in field- and laboratory-infected C. virginica has been known since the earliest reports of the parasite. Gener- ally, the parasite requires temperatures &20°C to multiply in oys- ters (Andrews 1988), but it can persist through the winter. Fisher et al. ( 1992) showed that elevated temperature increased the mor- tality of infected oysters. The direct correlation of disease preva- lence (and intensity) with increasing temperature has been shown in many laboratory studies (e.g., Chu and La Peyre 1993b, who also showed that total hemocyte count increased with higher tem- peratures [10-25°C]). Elevated total hemocyte counts in P. mari- /»/i-infected oysters at 25°C were noted, as seen during develop- ment of moderate-heavy infections by Anderson et al. (1992b). According to Chu and La Peyre ( 1993b), both lysozyme levels and condition index were negatively correlated with this disease, un- like the percent granulocytes, phagocytosis, and hemagglutinin levels. P. niariniis infection apparently did not alter temperature effects on the various defense-related activities measured; the pos- itive and negative correlations between these activities and tem- perature seen in uninfected oysters were identical in infected oys- ters. Ambient salinity is also reported to affect oyster immune pa- rameters as well as prevalence and intensity of P. marinus disease (La Peyre et al. 1989, Gauthier et al. 1990, Chu and La Peyre 1989, Chu et al. 1993). Increasing salinity suppresses hemocyte spreading and locomotion (Fisher and Newell 1986) and may af- fect hcmolymph lyso/ymc concentration, as described below. Low salinity has an inhibitory influence on P . niannns infection; transmission and progression can occur at 10 ppt (Chu and La Peyre 1993a). but sporulation is inhibited at 6 ppt (Chu and Greene 1989). P. nuinniis infection progression in oysters is retarded al 12 ppt and is stopped at «9 ppt (Ragone and Burreson 1993). Low salinity can delay or reduce oyster mortality caused by this disease (Scott et al. 1985, Ragone 1991). P. marinus-mi&cKcA oysters, when placed in low salinity water, have lower serum lysozyme levels and show increased survivorship (La Peyre et al. 1989, Ragone 1991. Chu et al. 1993). La Peyre et al. (1989) reported a weak correlation between total hemocyte count and salinity, but hemolymph total protein levels and hemagglutinin levels showed no relationship to salinity. The effect of salinity on lysozyme level in C. virginica is uncertain. For example, Chu and La Peyre (1989) found no correlation between salinity and lysozyme levels in a 1-year study but subsequently reported that lysozyme concen- trations decrease with elevated salinity (Chu et al. 1993). HEMOCYTIC DEFENSE AGAINST P. MARINUS: ROIs In mammalian leukocytes, appropriate membrane perturba- tions, caused by phorbol myristate acetate binding or phagocyto- sis, trigger the uptake of O,, generation of NADPH, activation of NADPH oxidase, and generation of superoxide anions (Oj). A number of other toxic ROIs such as hydrogen peroxide (H-,0,). singlet oxygen, hydroxyl radicals, hypochlorous acid, etc.. can be subsequently derived from 0~ (see reviews by Babior 1984; Kle- banoff 1985). Similar reactions seem to take place in molluscan hemocytes (Adema et al. 1991; Anderson 1994b). The ROIs are thought to play significant roles in phagocyte-mediated killing of microorganisms. They can also have deleterious effects when gen- erated at levels that exceed the antioxidant mechanisms available to the host. One of the most commonly used assays for the quantitation of ROIs is luminol-augniented CL, in which ROIs generated by the hemocytes activate the probe luminol. Upon relaxation, the lumi- nol molecules emit photons which can readily be measured in a luminometer or a liquid scintillation spectrometer adapted for sin- gle-photon counting. The CL activity of mammalian leukocytes has been directly linked to their antimicrobial activity (Horan et al. I Oyster Defense Responses 129 1982). This association of CL activity with defensive mechanisms has often been assumed in the case of molluscan hemocytes. Current data suggest that strong luminol-augmented CL is pro- duced when oyster hemocytes bmd and phagocytize yeast cells or zymosan particles (Larson et al. 1989. Fisher et al. 1990. Ander- son et al. 1992a). However, luminol-augmented CL induction by yeast or zymosan may not be seen in hemocytes of other bivalve species, such as Mercenaria merct'iuiriti (Cheng 1976a, Anderson 1994a). The CL assay with lummol actually measures a MPO- dependent antimicrobial pathway (DeChatelet et al. 1982. Dahl- gren and Stendahl 1983) in mammalian blood cells. MPO activity is present in bivalve hemocytes (Schlcnk et al. 19911, and treat- ment with specific MPO inhibitors will greatly reduce zymosan- stimulatcd, luminol-dependent CL of oyster hemocytes (Ander- son, unpublished). Therefore, oyster hemocytes probably have a mechanism analogous to the H,0,/MPO/halide antimicrobial sys- tem first characterized by Klebanoff (1968). What is the physiological significance of luminol-augmented CL to molluscan defensive capacities against infectious agents, especially P nuiriiuis? Several authors report that exposure to several classes of environmental contaminants can result in de- creased hemocyte CL responses (Larson et al. 1989, Fisher et al. 1990, Anderson et al. 1996. 1994. Coles et al, 1994), Exposure of oysters to chemical stressors can enhance latent and/or experi- mental P. marinus infections (Winstead and Couch 1988, Chu and Hale 1994, Anderson et al, 1996, Fisher et al. 1995). but it is very difficult to use CL to probe the underlying mechanism(s) involved. It is easy to dismiss studies of CL. or any other defense- related activity of C, virginica hemocytes. as irrelevant to un- derstanding this disease, because of the obvious inability of hemocytes to control ultimately the progression of the infection. However, there could be limited ROl-mediated anU-Perkiiisiis ca- pability involved in the elimination of the parasites during early stages of infection. There is electron microscopic evidence of lim- ited intrahemocytic P. marinus destruction (La Peyre 1993, Bushek et al. 1994). Perhaps the expression of this minimal ability to cope with initial, very light infections is important to determin- ing resistance. Eventually our understanding of the pathogenesis of this disease will be more complete, but the use of CL to fill in the details may be complicated. For example, it could be hard to interpret changes in zymosan- or yeast-induced, luminol-aug- mented CL in hemocytes from a study designed to evaluate the effects of an environmental contaminant on P. marinus progres- sion. As mentioned above, many chemical stressors will produce dose-dependent inhibition of CL, which could be considered the basis of immunosuppression and enhanced disease susceptibility. However, CL of C. virginica hemocytes has been shown to sig- nificantly increase with elevated levels of P. marinus infection (Anderson et al. 1995). Therefore, it is difficult to sort out the simultaneous and opposite effects of chemical stress and infection in an oyster experiencing both situations. In addition, some chem- icals that suppress CL at higher doses are actually stimulatory at low doses (Fisher et al. 1990, Anderson et al, 1994). The rela- tionship between high levels of P. marinus infection and increased CL is interesting in that it suggests a form of hemocytic activation produced by the intracellular presence of the parasite. We specu- lated that although this apparent activation did not provide effec- tive anti-P. marinus protection, it might serve to explain some of the pathogenic effects of the disease via increased ROI-mediated tissue damage (Anderson et al. 1992b). However, one cannot ex- trapolate data from zymosan-induced CL to those from P mari- «//i-induced CL. In fact, P. marinus ingestion seems to be inef- fective in stimulating CL in hemocytes withdrawn from oysters, regardless of the level of infection In the host (La Peyre et al. 1992. Anderson, unpublished). Clearly, the putative activation of hemocytes during P . mari- nus infection has little effect on the parasite in vivo, if zymosan is required to induce CL (ROl) generation. It is possible that phago- cytosis of many kinds of foreign particles (Including P. marinus) can prime or activate C . virginica hemocytes. but the expression of elevated CL responses depends on subsequent stimulation by specific agents (such as zymosan, but not P. marinus). If this is true, CL could prove to be a more useful mechanistic probe to study infectious diseases of C. virginica other than that caused by P. marinus. The inability of certain other bivalve parasites to elicit CL in zymosan-responsive hemocytes of their host has been pre- viously reported (Hervio et al. 1989a, b, LeGall et al. 1991 , Bach- ere et al. 1991 ), Leishmania and other parasites of vertebrates can enter host macrophages by interacting with receptors mediating internalization without triggering the respiratory burst (McNeely and Turco 1987, Russel and Talamus-Rohana 1989). CLASSIC OYSTER DEFENSE MOLECULES AND P. MARINUS Agglutinins The presence of lectins, or natural agglutinins, has been noted in many bivalves including oysters. Hemagglutinins can have op- sonic properties, as shown by their ability to enhance phagocytosis of foreign erythrocytes by hemocytes (McDade and Tripp 1967b). Bacterial agglutinins have also been described (Arimoto and Tripp 1977, Tamplin and Fisher 1989, Fisher and Dinuzzo 1991); in some cases these lectins can be induced by bacterial infection and may facilitate their immobilization and destruction by the host (Olafsen et al. 1992). Infection of C. virginica by Haplosparidium nelsoni (the causative agent of MSX) can produce alterations in the levels of serum agglutinins (Ling 1990. Chintala and Fisher 1991 ). For example. Vilirio choierae agglutinin titers were seasonally elevated in MSX-resistant strains of oysters (Chintala and Fisher 1991 ). In a study of P. marinus infections in oysters from the Gulf of Mexico held under different laboratory conditions, no associa- tion was found between antimammalian erythrocyte hemagglutinin levels and parasite infection intensity (Fisher et al. 1992). Re- cently Chintala et al. ( 1994) have studied the relationship of oyster serum agglutinins to the protozoan parasites P. marinus and H. nelsoni. No relationship was shown between initial baseline levels of hemagglutinins or bacterial agglutinins and postinfection sur- vival times of the oysters. Agglutinin levels were also not corre- lated with parasite densities during disease progression. Therefore, the lectins measured in this study probably play little or no role in defense reactions against these protozoan parasites. Lectins from various plant sources will bind to oyster hemo- cytes: concanavalin A has received considerable study in this re- gard (Yoshino et al. 1979). Recently, Cheng et al. (1993) reported that Latlnrus adoralus (sweet pea) lectin will bind and agglutinate C. virginica hemocytes. Subsequently, it was shown that hemo- cytes from P. manViwi-infected hosts were less subject to L. (idnratus lectin agglutination than cells from uninfected animals (Cheng and Dougherty 1994). This was explained by competition for the lectin by common saccharides on the extracellular parasites and the hemocytes. Sharing of these as yet unidentified saccha- rides was suggested to account for the failure of the hemocytes to 130 Anderson recognize and phagocytizc those meronts existing free in the se- rum. Lysozyme Oysters from an area of the James River with comparatively low salinity ( 10 ppt) showed higher P. mannus survival rates than those from higher salinity sites (20 and 32 ppt). The lysozyme concentration in the James River oysters' hemolymph was also higher than that seen in other groups, prompting the authors to suggest a causc-and-effect relationship (Chu and La Pevrc 19y3a. Chu et al. 1993). A positive correlation between lysozyme con- centration and C. virginua survival had been previously reported by La Peyre et al. (1989). Since both disease intensity and he- molymph lysozyme activity are negatively correlated with salinitv. the relative importance of lysozyme as a determinant in the out- come off. marinus infections is not clear. Its bactericidal activity is well documented, but its effect on P. marinus has yet to be directly assessed. Prophenoloxidase-Activating System The capability of phenoloxidase to mediate the conversion of phenolic substrates, such as i.-dopa. to melanin has been demon- strated in leukocytes from human beings and other vertebrates. The enzyme has also been detected in the hemocyies and he- molymph of many marine invertebrates (Smith and Soderhall 1991). Melanization occurs mainly in arthropods, where it plays a prominent role in many aspects of defense responses including non-self-rccognition, opsonization, cellular communication, anti- bacterial activity, wound healing, and encapsulation (Soderhall 1982, 1992). Phenoloxidase can be released by activation of the corresponding proenzyme by well-known immunostimulators such as pi,3 glucans (Pye 1974) and bacterial lipopolysaccharides (Soderhall 1982). A detailed description of phenoloxida.se activity in the hemo- cytes and serum of M. eiliilis (Coles and Pipe 1994) shows that this defense system can also be found in bivalves. The enzyme was localized in large granules of the eosinophilic hemocytes and showed significant seasonal variability. M. ediilis hemocytes showed phenoloxidase activation after preincubation with zvmo- san supernate, a treatment already known to cause a similar effect in crayfish hemocytes (Uneslam and Soderhall 1977). The precise role of phenoloxidase in bivalve immunity has not as yet been elucidated; however, it clearly merits more study, particularly in light of presently existing uncertainties about the significance of more commonly investigated putative immune mechanisms to de- fense against P . murinus and other pathogens. OTHER OYSTER- .AND P. .V/.4/f/A( S-ASSOCIATED FACTORS Proteases P. marinus infections inhibit growth in C. virginica (Paynter and Burreson 1991 ) and cause severe mortalities. Surprisingly, no major alterations in feeding rate, metabolic rate, and assimilation efficiency of C. virginica accompany heavy infection by P. mari- nus (Newell et al. 1994). Virulence factors associated with this parasite and mechanisms of pathogenicity seem to be poorly un- derstood at this time. A number of P. marnius-dcrived proteins can be detected in spent culture media, including proteases that could participate in necrotic reactions in infected oysters (Faisal et al. 1994, La Peyre and Faisal 199S). Some of these enzymes have been shown to be serine proteases and have been purified from P. marinus media by affinity chromatography (La Peyre et al. 1995). Five distinct bands could be eluted with molecular weights ranging from 35 to 55 kDa. Acid Phosphatase Acid phosphatase associated with parasites can play a role in avoidance of host defense reactions by disruption of phosphopro- teins and/or inhibition of superoxide anion production. Both P. marinus meronts and C. virginica hemocytes have been shown to contain acid phosphatase (Volety and Chu 1994). The intrahemo- cytic activity varied markedly depending on the geographic sources of the oysters, and both intrahemocytic and intrameront activities were positively correlated with ambient temperature. Living P marinus meronts significantly inhibited zymosan- stimulated oxyradical (CL) release by C. virginua hemocytes, whereas heat-killed P. marinus did not produce this effect (Volety and Chu 1995). Whether this is due in part to acid phosphatase- mediated ROI suppression by the viable parasites has yet to be determined. In this regard, it is interesting that estuarine water in which P. marinus had been maintained slightly suppressed zymo- san-stimulated CL: it was also shown to contain acid phosphatase, which was not detected in the appropriate control water. Charac- terization of extracellular products from P. marinus that suppress CL and other defense mechanisms of C. virginica will probably be an active area of research in the near future. Acid phosphatase is one such product that blocks O3 in leukocytes, permitting intra- cellular survival of parasites such as Leishmania (Remaley et al. 1984). as well as molluscan parasites such as Bonaima osireae in Ostrca ediilis (Hervio et al. 1991). Iron-Binding Proteins P. marinus has a strong requirement for soluble iron (Gauthier and Vasta 1994). Soluble iron in the culture medium enhances P. marinus replication, and the addition of natural iron chelators such as lactoferrin, transfenin, or desfcrrioxamine to the medium re- duces parasite proliferation in a dose-dependent manner. It was suggested that shifts in iron availability to the host and parasite during ongoing infection might affect the iron pools required for O5 and -OH production, which in turn might allow P. marinus to avoid intracellular oxidative damage. The characterization of the iron-binding proteins of oysters and their modulation during in- fection should provide valuable insight with regard to antimicro- bial defense mechanismis). Nonimmuni)logical mechanisms that involve simply withholding iron trom the pathogen may be im- portant in controlling nucrobial infections. Stress Proteins In ectothemiic animals like oysters the hemocytes must func- tion over a considerable temperature range; it has already been reported that environmental factors such as salinity and tempera- ture can influence certain hemocyte defense responses (Fisher 1988). As mentioned previously, increased mortality among oys- ters infected with P marinus is associated with high ambient temperature and salinity. A highly conserved biological response to hyperthermia, hypoxia, toxicants, and other noxious stimuli is the synthesis and accumulation of heat shock proteins and other stress proteins (SP). These proteins ensure the survival, normal functioning, and recovery of cells during and after stressful cir- cumstances. Among their numerous activities. SP have been shown to be important in the functioning of mammalian immune cells (DeNasiel and Pierce 1993. Youne et al. 1993). Manv SP Oyster Defense Responses 131 classes are present in aquatic organisms, including bivalves (Sand- ers 1993), but their possible role(s) in oyster hemocyte-parasite interactions are only recently being investigated. The level of the 70-kDa heat shock protein {SP70) in oyster hemocytes was reported to increase with increasing P. marinus infections (Brown et al. 1993). The SP70 level in mantle tissue was elevated during the late summer and fall, times when P. marinus infection is high. More recently the SP of both oyster hemocytes (Tirard et al. 1995a) and P. inanmis (Tirard et al. 1995b) have been characterized. Cold shock had no significant effect on protein synthesis by hemocytes, but heat shock produced by temperatures >20°C above ambient triggered synthesis of 32-, 34-, 37-, 70-, and 85-kDa SP. The hemocytes withstood the acute temperature increment well, since most remained intact and via- ble. It is possible that the induced SP played an important role in protecting key cellular proteins against denaturation, thus permit- ting the hemocytes to maintain their surveillance and effector func- tions during periods of thermal stress. P. marinus also will pro- duce heat shock proteins, but only at temperatures somewhat higher than those used to elicit SP in C. virginica hemocytes. The molecular masses of P. marinus SP were about 29, 63, 79, and 86 kDa, clearly distinguishable from those of the host. These differ- ences suggest the possibility of studying the SP responses of both host and parasite simultaneously in a mixed culture. The higher thermal threshold for the parasite's SP response may mean that it can retain normal function at temperatures that are very stressful to the hemocytes. Although the relevance of these observations to host-parasite interactions under environmental conditions encoun- tered in the field needs to be carefully evaluated, clearly this is an exciting area for future research. CONCLUSIONS AND FUTURE RESEARCH DIRECTIONS Based on our knowledge of host-parasite interactions, it is rea- sonable to speculate that the functional capacity of the immune system of C. virginica plays a pivotal role in the response elicited by the protozoan P. marinus. It has been suggested by several authors (e.g., Cheng 1987) that the virulence of P. marinus in- creases when its host becomes immunologically compromised; however, this conclusion is often based on indirect evidence. We are still limited by an insufficient knowledge of oyster defense mechanisms as related to P. marinus infections and/or incomplete understanding of the mechanisms underlying the pathogenicity of this parasite. The picture is further complicated by the difficulty of interpreting defense activities during progression of the disease. It is difficult to know to what extent the changes observed indicate factors determining resistance/susceptibility or merely represent responses to the presence of the parasite. In addition, a number of hemocyte activities can be modulated by environmental factors such as temperature and salinity; these same conditions can have profound direct effects on survival and multiplication of P. mari- nus. Perhaps the most direct link between C. virginica immunocom- petency and resistance to P. marinus has come from recent toxicity studies. In studies of the acute toxicity of the carcinogen n-nitro- sodiethylamine to oysters, this compound was found to enhance P. marinus infections (Winstead and Couch 1988). Oysters receiving sublethal doses developed locally heavy infections which triggered atypically light hemocytic responses, suggesting chemically in- duced immunosuppression relevant to P. marinus. In vitro expo- sures to other environmental contaminants, such as tributyltin (TBT), were known to reduce the ability of oyster cells to produce ROIs (Fisher et al. 1990). Subsequent studies have shown that in vivo TBT exposure of oysters will accelerate the progression of experimental P. marinus infections (Fisher et al. 1995, Anderson et al. 1996). In addition to these single-toxicant experiments, ex- posure of C. virginica to water-soluble fractions of polyaromatic hydrocarbon (PAH)-contaminated sediments enhanced preexisting P. marinus infections and increased susceptibility to experimental infection (Chu and Hale 1994). The above-mentioned data are interpreted by the investigators involved as likely examples of contaminant-induced reductions of oyster defensive capacity that are expressed as diminished resis- tance to latent or experimentally initiated P. marinus infections. However, the possibility that the chemicals produce direct stimu- latory effects on the parasite, or act on some nonimmunological parameter of disease resistance, is also raised. There seems to be little doubt that, under typical conditions, C. virginica has little ability to control the multiplication and progression of P. marinus once it becomes established in the hemolymph or other tissues. There is only circumstantial evidence that the best characterized putative antimicrobial defense mechanisms of C. virginica. such as lysozyme or ROIs, have significant deleterious effects on P. marinus. The question of the relative protective importance of these factors in other bivalve species less seriously affected by this parasite needs to be resolved. Clearly, the basis for the pathoge- nicity of P. marinus in minimally stressed C. virginica is poorly understood and the role of immunity in the process is hard to evaluate. However, since chemical and other stressors are able to accelerate the progression of this parasitic disease, perhaps xeno- biotic exposure-based models will contribute useful knowledge of physiological disease control mechanisms. In-depth studies of C. virginica:P. marinus interactions at the cellular and molecular levels are still needed. Such work should have interest to immu- nologists and parasitologists and will hopefully find practical ap- plication in managing and/or controlling this major disease of oys- ters. LITERATURE CITED Adema. C. M., W, P. W. van der Knaap & T. Sminia. 1991. Molluscan hemocyte-mediated cytotoxicity: the role of reactive oxygen interme- diates. Rev. Aqual. Sci. 4:201-223. Anderson, R. S. 1994a. Hemocyte-derived reactive oxygen imeniicdiale production in tour bivalve mollusks. Dev. Camp. Immunol. 18:89-96. Anderson, R. S. 1994h. Modulation of blood cell mediated oxyradical production in aquatic species: implications and applications. In: D. C. Malins & G. K. Ostranderleds.). Aquatic To.xicology: Molecular. Bio- chemical, and Cellular Perspectives . Lewis Publishers, Boca Raton, Florida, pp. 241-265. Anderson. R. S.. L. M. Oliver & D. Jacobs. 1992a, Immunotoxicity of cadmium for the eastern oyster (Crassosirea virginica [Gmelin, 1791 1): effects on hemocyte chemiluminescence. ,/, Slu'llfi.':h Res. 1 1: 31-.\S Anderson, R. S., K. T. Paynter & E. M. Burreson. 1992b. Increased reactive oxygen intermediate production by hemocytes withdrawn from Crassostrea virginica infected with Perkinsus marinus. Biol. Bull. 183:476-481. Anderson, R. S., K. T. Paynter & E M. Burreson 1992c. Infection of oysters with Perkinsus marinus enhances chemiluminescence by hemo- cytes. In vitro 28;82A. Anderson, R, S., L. M. Mora & S. A. Thomson. 1994, Modulation of 132 Anderson oyster (Crassostrea virginiiu) hemocyte immune function by copper, as measured by luniinol-enhanced chemiluminescence. Comp. Bio- chem. Physiol. 108C:2l5-220. Anderson. R. S., E. M. Burreson & K. T. Paynter. 1995. Defense re- sponses of liemocytes withdrawn from Cnissosrrea virginica infected witfi Perkinsiis muriniis. J. Inverlebr. Palhol. 66:82-89. Anderson. R. S.. M. A. Linger & E. M. Burreson. 1996. Enhancement of Perkinsiis marinus disease progression in TBT-exposed oysters (Crassoslrea virginica). Mar. Environ. Res. (in press). Andrews, J. D. 1988. Epizootiology of the disease caused by the oyster pathogen, Perkinsus marinus and its effects on the oyster industry Amer. Fish Soc. Spec. Publ. 18:47-63. Arimoto, R. & M. R. Tripp. 1977. Characterization of a bacterial agglu- tinin in the hemolymph of the hard clam Mercenaria mercenaria J Imeriebr. Palhol. .30:406-413. Babior, B. M. 1984. O.xidants from phagocytes: agents of defense and destruction. Blood 64:959-966. Bachere, E., D. Hervio & E. Mialhe. 1991. Luminol-dependent chemilu- minescence by hemocytes of two marine bivalves, Osirea eclulis and Crassostrea gigas. Dis. Aquat. Org. 1 1:173-180. Brown, D., B, P. Bradley & K. T. Paynter. 1993. The physiological effects of protozoan parasitism on the eastern oyster. Crassostrea vir- ginica: induction of stress protems, J. Shellfish Res. 12:135-136. Bushek, D.. S. K. Allen, K. A. Alco.x. R. Gustafson & S. Ford. 1994. Dose response of the eastern oyster, Crassostrea virginica. to cultured cells of Perkinsiis marinus. the agent of Demio disease. J. Shellfish Res. 13:313. Cheng, T. C. 1976a. Aspects of substrate utilization and energy require- ment during molluscan phagocytosis. J. Invenebr. Palhol. ll.ld'i- 268. Cheng, T- C. 1976b. Beta-glucuronidase in the serum and hemolymph cells of Mercenaria mercenaria and Crassostrea virginica (Mollusca: Pelecypoda). J. Imertebr. Pathol. 27:125-128. Cheng, T. C. 1978. The role of lysosomal hydrolases in molluscan cellular responses to immunologic challenge. Comp. Pathohiol. 4:59-71. Cheng, T. C. 1981. Bivalves. In. N. A. Ratcliffe & A. F. Rowley (eds.). Invertebrate Blood Cells, Vol. I. Academic Press, London, pp. 233- 300. Cheng, T. C. 1987, Some cellular mechanisms governing self and nonself recognition and pathogenicity in vertebrates and invertebrates relative to protistan parasites. Aqiiacultiire 67:1-14. Cheng, T. C. 1988. In vivo effects of heavy metals on cellular defense mechanisms of Crassostrea virginica: total and differential cell counts. J. Imertebr. Palhol. 51:207-214. Cheng. T. C. 1992. Selective induction of release of hydrolases from Crassostrea virginica hemocytes by certain bacteria. J Imertebr. Pathol. 59:197-200. Cheng, T. C. & W. J. Dougherty. 1994. Inhibition of hemocyte aggluti- nation by Lathyriis odoratiis lectin in Chesapeake Bay, USA, oysters (Crassostrea virginica) infected with Perkinsiis marinus. Res. Rev. Parasitol. 54:121-123. Cheng, T. C. & G. E. Rodrick. 1974. Identification and characterization of lysozyme from the hemolymph of the soft-shelled clam Mya are- naria. Biol. Bull. 147:311-320. Cheng, T. C. & G. E. Rodrick. 1975. Lysosomal and other enzymes in the hemolymph of Crassostrea virginica and Mercenaria mercenaria. Comp. Biochem. Physiol. 52B:443-447. Cheng. T. C, G. E. Rodrick, D. A. Foley & S. A. Koehler. 1975. Re- lease of lysozyme from hemolymph cells of Mercenaria mercenaria during phagocytosis. J. Imertebr. Pathol. 25:261-265. Cheng, T. C, W. J. Dougherty & J. Burrell V. G 1993. Lectin-binding differences on hemocytes of two geographic strains of the American oyster, Crassostrea virginica. Trans. Am. Microsc. Soc. 1 12:151-157. Chintala, M. M. Sc W. S. Fisher. 1991. Disease incidence and potential mechanisms of defense for MSX-resistanl and susceptible eastern oys- ters held in Chesapeake Bay. J. Shellfish Res. 10:439-443. Chintala, M. M., S. E. Ford. W. S. Fisher & K. A. Ashton-Alcox. 1994. Oyster serum agglutinins and resistance to protozoan parasites. J. Shellfish Res. 13:115-121. Chu, F.-L. E. & K. H. Greene. 1989. Effect of temperature and salinity on in vitro culture of the oyster pathogen, Perkinsus marinus (Api- complexa: Perkinsea). J. Imertebr. Pathol. 53:260-268. Chu. F -L E. & R. C. Hale. 1994. Relationship between pollution and susceptibility to infectious disease in the Eastern oyster. Crassostrea virginica. Mar. Environ. Res. 38:24,V256. Chu, F.-L. E. & J. F. La Peyre. 1989. Effect of environmental factors and parasitism on hemolymph lysozyme and protein of American oysters (Crassostrea virginica). J. Invertebr. Pathol. 54:224—232. Chu. F.-L. E. &J. F. La Peyre. 1993a. Development of disease caused by the parasite, Perkinsus marinus and defense-related hemolymph factors in three populations of oysters from the Chesapeake Bay, USA. J. Shellfi.'^hRes. 12:21-27. Chu, F.-L. E. & J. F. La Peyre. 1993b. Perkinsus marinus susceptibility and defense related activities in eastern oysters Crassostrea virginica: temperature effects. Dis. Aquat. Org. 16:223-234. Chu, F -L , J. F. La Peyre & C. S. Burreson. 1993. Perkinsus marinus infection and potential defense-related activities in Eastern oysters, Crassostrea virginica: salinity effects. J, Invertebr. Pathol. 62:226- 232. Coles, J- A- & R. K. Pipe. 1994. Phenoloxidase activity in the haemo- lymph and haemocytes of the marine mussel Myiiliis ediilis. Fish Shell- fish Immunol. 4:337-352. Coles, J. A., S. R. Farley & R, K, Pipe. 1994. Effects of fluoranthene on the immunocompetence of the common marine mussel. Mytilus edulis. Aquat. Toxicol. 30:367-379. Dahlgren, C. & O. Stendahl. 1983. Role of myeloperoxidase in luminol- dependent chemiluminescence of polymorphonuclear leukocytes. In- fict. Immiin. 39:736-741. DeChatelet, L R.. G. D, Long. P. S. Shirley, D, A, Bass, M. J. Thom- as, F. W. Henderson & M. S. Cohen. 1982. Mechanism of the lumi- nol-dependent chemiluminescence of human neutrophils J Immunol. 129:1589-1593 DeNagel, D. C. & S. K. Pierce. 1993. Heat shock proteins m immune responses. Crit. Rev. Immunol. 13:71-81. Faisal, M.. J. F. La Peyre, D. Y. Schafliauser & E. H. Rizkalla. 1994. Detection of proteases in the supemates of Perkinsus marinus cultures. J. Shellfi.'ih Res. 13:296. Feng, S. Y. 1965a. Heart rate and leucocyte circulation in Crassostrea virginica. Biol. Bull. 128:198-210. Feng, S. Y. 1965b. Pinocytosis of proteins by oyster leucocytes. Biol Bull. 129:95-105. Feng, S. Y. & W. J. Canzonier. 1970. Humoral responses in the Amer- ican oyster (Crassostrea virginica) infected with Bucephalus sp, and Minchinia nelsoni. Spec. Publ. Amer. Fish. Soc. 5:497-510. Feng. S. Y. & J. S. Feng. 1974. The effect of temperature on cellular reactions of Crassostrea virginica to injections of avian erythrocytes. J. Invertebr. Pathol. 23:22-37. Fisher, W. S. 1986. Structure and function of oyster hemocytes. In: M. Brehelin (ed. ). Immunity in Invertebrates Springer- Verlag, Berlin, pp. 25-35. Fisher, W S. 1988. Environmental influence on bivalve hemocyte func- tion. Amer. Fish. Soc. Spec. Publ. 18:225-237. Fisher, W S & A. Dinuzzo. 1991. Agglutination of bacteria and eryth- rocytes by serum from six species of marine molluscs. J. Invertebr. Pathol. 57:380-394. Fisher, W. S. & R. I. E. Newell. 1986. Salinity effects on the activity of granular hemocytes of American oysters, Crassostrea virginica. Biol Bull. 170:122-134. Fisher. W. S. & M. Tamplin. 1988. Environmental intluence on activities and foreign-particle binding by hemocytes of American oysters, Cras- sostrea virginica. Can. J. Fish Aquat. Sci. 45:1309-1315. Fisher. W. S., A. Wishkovsky & F.-L. E. Chu. 1990. Effects of tri- butyltin on defense-related activities of oyster hemocytes. Arch. Envi- ron. Contain. To.xicol. 19:354-360. Oyster Defense Responses 133 Fisher. W. S., J. D. Gaulhier & J, T. Winstead. 1992. Infection intensily oi Perkinsiis marimis disease m Crassoslrea virginica (Gmelin. 1791) from the Gulf of Mexico maintained under different laboratory condi- tions. J. Shflllish Res. 11:263-269. Fisher. W. S.. J. T, Winstead. L. M. Oliver & P. E. Edwards. I99.V Physiological and immunological measures of Apalachicola Bay oys- ters during a one-year period. 7. Shellfish. Res. 12:127. Fisher, W. S.. L. M. Oliver. E. B. Sutton. C. S. Manning & W. W. Walker. 1995. Exposure of eastern oysters to tributyltin increases the severity of Perkinsus marimis disease. J. Shellfish Res. 14:265-266. Foley, D. A. & T. C. Cheng. 1975. A quantitative study of phagocytosis by hemolyniph cells of the Pelecypods Cnissoslreu virginicci and Mer- ceiuina mercenaria. J. Imeriehr. Palhol. 25:189-197. Foley. D, A. & T. C. Cheng. 1977. Degranulation and other changes of molluscan granulocytes associated with phagocytosis. J. Imeriehr. Palhol. 29:321-325. Fries. C. R. 1984. Protein hemolyniph factors and their roles in inverte- brate defense mechanisms: a review. In: T. C. Cheng (ed.). Compar- ative Pathohioloiiy. Vol. 6. Plenum. New York. pp. 49-109. Gauthier. J. D. & G. R. Vasta. 1994 Inhibition of in vitro replication of the oyster parasite Perkinsus nutrinus by the natural iron chelators transfemn. lactoferrin and desfenrioxamine. Dew Comp. Immunol. 18: 277-286. Gauthier. J. D.. T. M. Soniat & J, S Rogers 1990. A parasitological survey of oysters along salinity gradients in coastal Louisiana, J . World Aquacuh. Soc. 21:10.5-115. Hardy. S. W.. T. C. Fletcher & J. A. Olafsen. 1977. Aspects of cellular and humoral defense mechanisms in the Pacific oyster. Crassoslrea .i>ii;as In: 1. B. Solomon & J. D. Horton (eds.). Developmenlal Im- munohioloi>\ . Elsevier. Amsterdam, pp. 59-66. Hervio. D.. E. Bachere, E. Miaihe & H, Grizel. 1989a. Chemiluminescent responses of Oslrea etlulis and Crassoslrea gigas to Bonamia oslrea (Ascetosporal. Der. Comp. Immunol. 13:449. Hervio. D.. E. Bachere. E. Miaihe & H. Grizel. 1989b. Zymosan and specific-rickettsia activation of oxygen free radicals production in Peitcn ma.\imiis hemocytes. Dew Comp. Immunol. 13:448. Hervio. D.. D. Chagot. P. Godin. H. Grizel & E. Miaihe. 1991. Local- ization and characterization of acid phosphatase activity in Bonamia oslreae (Ascetosporal, an intrahemocytic protozoan parasite of the Hat oyster Oslrea etlulis (Bivalvial. Dis. Aqual. Org. 12:67-70, Horan. T, D,. D, English & T. A. McPherson. 1982. Association of neu- trophil chemiluminescence with microbicidal activity. Clin. Immunol. Immunopalhol. 22:259-269. Klebanoff, S. J. 1968. Myeloperoxidase-halide-hydrogen peroxidase an- tibacterial system. J. Bacleriol. 95:2131-2138, Klebanoff, S. J. 1985. Oxygen-dependent antimicrobial systems of mono- nuclear phagocytes. Prog. Leukocyte Biol. 4:487-503, La Peyre. J. F. 1993, Studies on the oyster pathogen Perkinsus marimis (Apicomplexa): interactions with host defenses of C. virginica and C gigas. and in vitro propagation [Ph.D. Dissertation], The College of William and Mary. Gloucester Point. Virginia. La Peyre. J. F. & M. Faisal, 1995. Perkinsus marimis produces extracel- lular proteolytic factor(s) in vitro. Bull. Eur. Assoc. Fish Palhol. 15: 28-3 1 , La Peyre. J. F.. F.-L. E. Chu & L. M. Ragone, 1989. The effect of salinity and Perkinsus marinus infection on some immunological pa- rameters of the oyster Crassoslrea virginica. J . Shellfish Res. 8:469. La Peyre. J. F., F.-L. E. Chu & W. K. Vogelbein. 1992. In vitro inter- action of Perkinsus marinus with hemocytes from eastern and Pacific oysters. Crassoslrea virginica and Crassoslrea gigas. J Shellfish Res. 11:200. La Peyre. J. F.. D, Y. SchatTiauscr. E. M. Rizkalla & M, Faisal. 1995. Production of serine proteases by the oyster pathogen Perkinsus mari- nus (Apicomplexa) in vitro. J. Euk. Microbiol. 42:451-458. Larson. K. G.. B. S. Roberson & F. M. Helnck. 1989. Effect of envi- ronmental pollutants on the chemiluminescence of hemocytes from the American oyster Crassoslrea virginica. Dis. Aqual. Org. 6:131-136. LeGall. G.. E. Bachere & E. Miaihe. 1991. Chemiluminescence analysis of the activity of Pecien maximus hemocytes stimulated with zymosan and host-specific Rickettsiales-like organisms. Dis. Aqual. Org. 11: 181-186. Ling, W. J. 1990. Cellular and humoral responses of resistant and sus- ceptible oysters, Crassoslrea virginica. to the infection of Haplo- sporidium nelsoni (MSX) |M.S. Thesis]. University of Connecticut, Storrs. Connecticut. McDade, J. E. & M. R. Tripp. 1967a. Lysozyme in the hemolymph of the oyster, Crassoslrea virginica. J. Inverlebr. Pathol. 9:531-535, McDade, J. E. & M. R.Tripp. 1967b. Mechanism of agglutination of red blood cells by oyster hemolyniph. J. Inverlebr. Palhol. 9:523-530. McNeely, T. B. & S. J. Turco. 1987, Inhibition of protein kinase C by Leishmania donavani lypophosphoglycan, Biochem. Bioplns. Res. Commun. 148:653-657, Mullainadhan, P, & L, Renwrantz, 1986. Lectin-dependent recognition of foreign cells by hemocytes of the mussel, Mxtilus edulis. Immunobi- ologx I7I:26-V273. Newell, R. 1. E., K. T. Paynter & E. M. Burreson. 1994. Physiological effects of protozoan parasitism of the Eastern oyster Crassoslrea vir- ginica: feeding and metabolism. J. Shellfish Res. 13:294. Olafsen, J. A., T. C. Fletcher & P. T. Grant. 1992. Agglutinin activity in Pacific oyster (.Crassoslrea gigas) hemolyniph following in vivo Vibrio anguillarum challenge. Dev. Comp. Immunol. 16:123-138. Oubella. R.. P. Maes. C, Paillard & M. Auffret. 1993. Experimentally induced variation in hemocyte density for Ruditapes philippinarum and R dccu.ssaius (Mollusca. Bivalva). Dis. Aqual. Org. 15:193-197. Paynter. K. T. & E. M. Burreson. 1991. Effects of Perkinsus marinus infection in the Eastern oyster. Crassoslrea virginica: II. Disease de- velopment and impact on growth rate at different salinities. J. Shellfish Res. 10.425^31. Pickwell. G. V, & S. A. Stcinert. 1984. Serum biochemical and cellular responses to experimental cupric ion challenge in mussels. Mar. En- viron. Res. 14:245-265. Pipe. R. K. 1990. Hydrolytic enzymes associated with the granular haemocytes of the marine mussel Mytilus edulis. Hislochem. J. 22: 595-603. Pye. A, E, 1974, Microbial activation of prophenoloxidase from immune insect larvae. Nature 251:610. Ragone. L M. 1991. The effect of low salinity on established infections of Perkinsus marinus (Apicomplexa: Perkinsasica) in the eastern oys- ter. Crassoslrea virginica ]M.S. Thesis). College of William and Mary, Gloucester Point, Virginia. 52 pp. Ragone. L. M. & E. M. Burreson. 1993. Effect of salinity on infection progression and pathogenicity of Perkinsus marinus in the Eastern oyster. Crassoslrea virginica (Gmelin). J. Shellfish Res. 12:1-7. Remaley, A T , D B. Kuhns, R E. Basford. G. H. Glew & S. S. Ka- plan 1984. Leishmanial phosphatase blocks neutrophil O'" produc- tion J Biol. Chem. 259: II 173-1 1 175 Renwrantz. L. 1983. Involvement of agglutinins (lectins) in invertebrate defense reactions: the immunobiological importance of carbohydrate- specific binding molecules. Dev. Comp. Immunol. 7:603-608. Renwrantz. L. 1990, Internal defense system of Mytilus edulis. In: G. B. Stefano (ed. I. Studies in Neuroscience: Neurobiology o/ (Mytilus edu- lis), Manchester University Press. Manchester, pp. 256-275. Renwrantz. L. & A. Stahmer. 1983. Opsonizing properties of an isolated hemolymph agglutinin and demonstration of lectin-like recognition molecules at the surface of hemocytes from Mytilus edulis. J. Comp. Phvsiol. 149:535-546. Russel. D. G. & P. Talamus-Rohana. 1989. Leishmania and the macro- phage: a marriage of niisconvenience. Immunol. Today 10:328-333. Sanders. B, M, 1993. Stress proteins In aquatic organisms: an environ- mental perspective. Crit. Rev. To.xicol. 23:49-75. Schlenk. D,. P. Garcia Martinez & D. R. Livingstone. 1991. Studies on myeloperoxidase activity in the common mussel. Mytilus edulis L. Comp. Biochem. Phvsiol. 99C:63-68. Scott. G. I.. D. P. Middaugh & T. I. Sammons. 1985, Interactions of 134 Anderson chlonne-produced oxidants (CPO) and salinity in affecting lethal and sublethal effects in the eastern or American oyster, Crassosirea vir- ginica (Gmelin), infected with the protistan parasite. Perkinsus man- nus. In: F. F. Vemberg, F. P. Thurgerg, A Calabrese. & W B Vemberg. Eds. Murine PoUurion and Physioloiiy: Recent Advances. University of South Carolina Press. Columbia. South Carolina, pp. 351-376. Smith. V. J. & K. Soderhall. 1491. A comparison of phenoloxidase ac- tivity in the blood of marmc invertebrates. Dev. Camp Immunol. 15:251-261. Soderhall, K. 1982. The prophenoloxidase activating system and nielani- sation: a recognition mechanism of arthropods? — A review. Dev. Comp. Immunol. 6:601-61 1 Soderhall, K. 1992. Biochemical and molecular aspects of cellular com- munication in arthropods. Ball. Zool. 59:141-151. Steinert. S. A. & G. V. Pickwell. 19X4. Quantitative determination of lysozyme in the hemolymph of MylUii.s ednlis by the enzyme-linked immunosorbent assay (ELISAl. Mar. Environ. Res. 14:229-243. Tamplin, M. L. & W. S. Fisher. 1989. Occurrence and characteristics of agglutination of Vibrio cholerae by serum of the eastern oyster, Cras- sosirea virginica. Appl. Environ. Microbiol 55:2282-2287. Tirard, C. T., R. M. Grossfeld, J. F. Levine & S. Kennedy-Stoskopf 1995a. Effect of hyperthermia in vitro on stress protein synthesis and accumulation in oyster haemocytes. Eisli Shellfish Immunol. 5:9-25. Tirard, C. T., R. M. Grossfeld, A. K. Volety & F.-L. E. Chu. 1995b. Heat shock proteins of the oyster parasite Perkinsus marinus. Dis Aqual. Org. 22:147-151. Tripp, M. R. 1966. Hemagglutinin in the blood of the oyster. J . Inveriebr. Pathol. 8:478-484. Unestam, T. & K. Soderhall. 1977. Soluble fragments from fungal cell walls elicit defense reactions in crayfish. Nature 267:45. Vasta, G. R. 1991. The multiple biological roles of invertebrate lectins: their participation in nonself recognition mechanisms, //i; G. W. Warr & N. Cohen (eds.). Phylogenesis of Immune Funclums. CRC Press, Boca Raton, Florida, pp. 73-1(11 Vasta, G. R., J. T. Sullivan, T. C. Cheng, J. J. Marchaloms & G. W. Warr. 1982. A cell membrane-associated lectin of the oyster hemocyte. J. Invenehr. Pathol. ■iO-.}^!-}!! . Vasta, G. R., T. C. Cheng & J. J. Marchalonia. 1984. A lectm on the hemocyte membrane of the oyster iCrassoslrea virginica). Cell. Im- munol. 99:475-488. Volety, A. K. & F.-L. E. Chu, 1994, A comparative study of acid phos- phatase in the parasite Perkinsus inarinus and lis host Crassosirea virginica. J Shellfish Res 13:297. Volety, A. K. & F,-L. E, Chu, 1995, Suppression of chcmiluminesccnce of eastern oyster (Crassosirea virginica) hemocytes by the protozoan parasite Perkinsus marinus. Dev. Comp. Immunol. 19:135-142, Winstead. J. T. & J. A. Couch. 1988. Enhancement of protozoan patho- gen Perkinsus marinus infections in American oysters Crassosirea vir- ginica exposed to the chemical carcinogen n-nitrosodielhylamine (DENA). Dis. Aqual. Org. 5:205-213 Yoshino, T. P. & T. C. Cheng. 1976. Experimental induced elevation ot aminopeptidase activity in hemolymph cells of the American oyster, Crassosirea virginica J. Inveriebr. Palhol. 11:261-110. Yoshino, T, P . L, R Rcnwrantz & T. C, Cheng. 1979, Binding and redistribution of surface membrane receptors of concanavatin A on oyster hemocytes. J. E.xp. Zool. 207:439-449. Young, D,. E. Roman. C Moreno. R OBnen & W, Bom, 1993 Mo- lecular chaperones and the immune response Philos Trans. R. Soc. Lond. B Biol. Sci. 339B:36.V3h7 jKiinml ,>f SlwUfish Research. Vol, 15. No. 1. 135-140. 19%. GENETIC ASPECTS OF DISEASE RESISTANCE IN OYSTERS PATRICK M. GAFFNEY' AND DAYID BUSHEK" ^ College of Marine Studies Universit}' of Delaware Lewes. Delaware 19958 'Belle Bariuh Institute for Marine Biology and Coastal Research University of South Carolina Georgetown, South Carolina 29442 ABSTRACT Like other phenotypic traits, resistance to disease is generally subject to underlying genotypic variability. This may come about indirectly, as a result of variation in overall physiology or in life history characteristics, or may be more directly attributable 10 variation in cellular or biochemical mechanisms. We summarize here our current understanding of genetic mtlucnccs on physio- logical and life history variation in Crassoslrea and review the evidence available to date on intra- and interspecific genetic variation in disease resistance, with emphasis on Perkinsiis marinus and MSX. We also describe our current view of population structure in Crassostrea virgiiiica, and how it may affect the evolution of disease resistance. Finally, we explore approaches to the development of disease-resistant oysters that capitalize on the genetic variability inherent in C viri;iiticu and within the genus Crassostrea. KEY WORDS: Crassostrea virginlca. Crassostrea gigas. Perkinsus nmriiim. Haptosporidium nehoni. MSX. disease, polyploidy, genetics, heterozygosity, physiology, population structure INTRODUCTION The resurgence of infectious human diseases in modern times serves as a sobering reminder that host-pathogen relationships are rarely static; rather, they are dynamic on both ecological and ev- olutionary time scales. This view is colorfully expressed by the ■■Red Queen hypothesis" of Van Valcn (1973). in which both parties are engaged in an incessant evolutionary arms race. (In Lewis Carroll's Through the Looking Glass, the Red Queen said to Alice, ■'here, you see, it takes all the running you can do to keep in the same place.") In this chapter, we focus on the role that genetic variability in the oyster host may play in determining its response to Perkinsiis itiarinus. Physiological Variation Physiological variation in the oyster may influence both the acquisition of disease and its subsequent development. For exam- ple, ventilation and clearance rates in bivalves vary considerably in response to a number of environmental factors such as season, salinity, temperature, nutrition, and turbidity; they also vary dur- ing the course of development and with respect to body size. If the likelihood of infection is a function of the rate of exposure to infective particles, then any physiological variation affecting pumping and clearance rates may alter disease risk. It also seems likely that physiological status will affect the progression of dis- ease following infection, although it is difficult in this case to separate cause and effect. Certainly a large proportion of physiological variation in nat- ural populations may be attributed to local environmental influ- ences, as demonstrated by reciprocal transplant experiments (Wid- dows et al. 1984). It seems likely that a substantial fraction of differences in disease resistance among populations that occupy geographically proximate but physically different environinents (e.g.. Chu and La Peyre 1993) may represent physiological accli- mation rather than genetic differences, but definitive data are lack- ing. At the same time, physiological functions may also be influ- enced by genotype. For example, standard metabolic rate (mea- sured as oxygen consumption by nonfeeding animals) in Crassos- trea virginica was found to be inversely related to heterozygosity at a set of five enzyme-coding (allozyme) loci (Koehn and Shum- way 1982). In a similar vein, more heterozygous oysters appeared to lose weight less rapidly when starved (Rodhouse and Gaffney 1984). Other studies have failed to detect associations between allozyme heterozygosity and physiological traits, and it remains unclear whether the association is genuinely sporadic or simply weak and therefore difficult to detect consistently (Gaffney 1986, 1990). Studies of the mussel Mylilus edulis demonstrate a positive relationship between multilocus allozyme heterozygosity and fit- ness that appears to result from lower protein turnover (and thereby a lower energetic demand for maintenance metabolisin) in more heterozygous individuals (Hawkins et al. 1989). Although this suggests a metabolic basis for allozyme-associated heterosis in bivalves, the mechanism(s) responsible for the connection be- tween genotype and physiology is still unknown. Regardless of the actual mechanism, genotypic variation in whole-body metabolism may translate into genotypic variation in resistance to pathogens, if the latter depends on net energy balance or rate of protein turn- over. In general, physiological variation, whether due to environ- mental influences, heredity, or both, may result in differences in susceptibility to infection. For example, individuals with higher filtration rates may enjoy an energetic advantage from greater food consumption, but at the same time could experience increased risk of exposure to infection with P. martinis. We are unaware, how- ever, of any experimental data testing this hypothesis and note that the stress of spawning does not appear to accelerate the develop- ment of MSX infections in C . virginica (Ford and Figueras 1988). Likewise, P. marinus infection intensity appears to be independent of sex and reproductive stage (Burrell et al. 1984. Wilson et al. 1990). Response to selection is a traditional method of demonstrating genetic variability in a trait. Although several studies have docu- mented differences in resistance to MSX between wild oysters and selected lines (Haskin and Ford 1988, Ewart et al. 1988, Hawes et 135 136 Gaffney and Bushek al. 1990. Matthiessen et al. 1990. Paynter and DiMichele 1990). no detailed physiological investigations have been conducted on selected lines. Life History Variation Variation in life history traits may be regarded as a higher order expression of physiological variation, in which differences in in- tegrated physiological functions lead to variability in features of primary ecological importance, e.g.. growth rate, fecundity, age at sexual maturation, or tuning of reproduction. Two types of evidence point to considerable genetic variability in life history traits in the oyster. Numerous studies have demonstrated a positive correlation be- tween heterozygosity at allozyme loci and growth rate in natural oyster populations (reviewed by Zouros and Foltz 1987). One might anticipate that greater somatic growth would also result in greater fecundity, given the generally positive relationship be- tween body size and fecundity. Although this has not been exam- ined in oysters, a study of A/. ediiUs showed that more heterozy- gous individuals tended to have greater fecundities (Rodhouse et al. 1986). Less is known about the relationship between heterozy- gosity and viability. Heterozygosity was positively associated with survival in C. viri^inka (Zouros et al. 198.'?) but negatively asso- ciated with survival in Osirea ediilis (Alvarez et al. 1989). Studies of other bivalve species generally point to positive associations (e.g.. Blot and Thiriot-Quievreux 1989. Borsa et al. 1992. Pecon Slattery et al. 1993). but sometimes no relationship is apparent (e.g.. Gaffney 1990. Fevolden 1992). Although the physiological correlates of increased heterozy- gosity are generally thought to result in enhanced fitness, this need not always be so. In the flat oyster C. editlis. a negative correlation between allozyme heterozygosity and viability was attributed to the enhanced susceptibility of larger (more heterozygous) oysters to the pathogen Bomimia ostreae (Alvarez et al. 1989). However, no direct evidence on the relationship between host genotype and rates of parasitic infection was available. A second, more indirect indication of genetic variation in life history traits is the existence of considerable latitudinal variation in growth rate and reproductive cycles. Although the majority of this variation may be environmental in origin, it appears that at least some is genetic. Barber et al. ( 1991 ) found that a hatchery stock of C. virginica derived from Long Island Sound initiated gonadal development and spawned 1 month earlier than native (Delaware Bay) controls, even after 5-6 generations of rearing in Delaware Bay. Hatchery lines of Virginia origin likewise spawn later than Delaware Bay lines (Ford et al. 1990). The relationship between life history traits and susceptibility to disease in oy.sters is purely conjectural at this point. However, given that gametogenesis and spawning arc major physiological events, requiring the mobilization of energy reserves and imposing a substantial physiological load, it seems plausible that they may alter the oyster's susceptibility to infection. It is also possible that biochemical or metabolic changes associated with reproduction may affect the process of infection or disease development. Avail- able data on this point are discussed below. Variation in life history parameters may also represent an ev- olutionary adaptation to disease pressure. High disease-induced mortality may favor reproduction at an earlier age. before the onset of disease (or before its effects become debilitating). Sexual mat- uration has been observed to occur in C . virgimca at 3 months of age in New Jersey (Bushek and Allen, pers. comm.) and at I month of age in Gulf Coast oysters (Hopkins 1954). indicating that early reproduction is a feasible strategy. However, there is no evidence that oysters have taken this course. In the absence of solid data on population structure and demography in the eastern oyster, we can only speculate on the factors responsible for the apparently limited ability of natural populations of C . virginica to develop resistance to P. munnus and MSX. Intra- and Interspecific Variation Any species that possesses an adequate store of genetic vari- ability can be expected to exhibit genotypic variation in response to a pathogen. Differences in susceptibility may result from overall physiological or metabolic variation, as discussed above, or from more specific biochemical or cellular responses. Genetic variation may exist at several levels, affecting the probability of exposure to pathogens, the uptake of pathogens in contact with tissue surfaces, and the progression of disease following infection. Different species may or may not differ considerably with re- spect to the same traits that vary within species; often interspecific differences may be far greater than intraspecific ones. This truism is central to agricultural breeding programs, in which within- species selection is often supplemented by efforts to move genetic elements from one species to another. Currently available data (presented below) suggest that this is also true for Crassostrea. EVIDENCE FOR GENETIC VARIATION IN DISEASE RESISTANCE IN CRASSOSTREA To date, studies of genetic variation in disease resistance in oysters have involved direct challenges of candidate oysters with pathogens, cither in field or laboratory settings, followed by as- says of disease prevalence and intensity. Studies of intraspecific variation have focused primarily on evaluations of hatchery strains, whereas interspecific comparisons have generally com- pared the Pacific oyster {Crassostrea gigas) to wild or susceptible hatchery populations of C. virginica. In addition, intraspecific genetic variation in disease susceptibility is indirectly demon- strated by the evolution of resistance in disease-challenged natural populations. The rate of evolution may be modest in some cases (e.g., resistance to MSX) and dramatic in others (e.g.. rapid evo- lution of resistance to Malpeque Bay Disease (Logic et al. I960)). However, the extent to which this apparent evolution represents genetic change in the parasite vs. genetic change the host is usually unknown. Intraspecific Variation Clearly, there is evidence for genetic variation in physiology and life history traits in C. virginica. It is not clear, however, to what extent this is matched by variation in the ability of the Amer- ican oyster to resist or tolerate infection with P. niariniis. Because the study of intraspecific variation in oyster resistance to P. marinus has been limited, it may be instructive to consider the better-studied case of MSX {Haplosporidium nelsoni). De- cades of selective breeding of C. virginica for resistance to MSX have resulted in several oyster lines that show a markedly en- hanced ability to tolerate infection with this pathogen (Haskin and Ford 1979). This response to selection argues strongly for the existence of useful genetic variation within populations. At the same time, the fact that tolerance is not complete, and that the ability to actually resist (rather than merely tolerate) infection has Genetic Aspects of Disease Resistance 137 not noticeably increased, suggests that the oyster's repertoire of genetic variability is not sufficient to provide complete protection against MSX. It is also worth noting that MSX-resistant oyster lines do not appear to have enhanced resistance to P. mariiiiis (Chintala and Fisher 14S9). Although parasitic infection appears to exert manifold physio- logical effects on its host, not all physiological variation in the host directly affects susceptibility to parasites. In a common-garden experiment. Ford et al. ( IWO) noted that the timing and intensity of MSX Infection were comparable for three strains of C. virgi- nica. despite marked differences in their reproductive state. Con- versely, Barber et al. (1991) found that selected and unselected oysters differed dramatically in degree of susceptibility to MSX but showed no detectable differences in metabolic rates prior to infection. Unfortunately, it appears unlikely that genetically mediated resistance to MSX will also confer resistance to P. inanims: ge- netic variation in resistance to one pathogen often has little bearing on resistance to other pathogens. In rainbow trout, artificial selec- tion for high and low stress response (as measured by blood Cor- tisol levels) was conducted, in the hope that trout with lower stress response would prove generally less susceptible to disease. Trout selected for low stress response showed lower mortality than a high stress response line when challenged with Aerainonus. but significantly higher mortality when infected with Vibrio (Fevolden et al. 1992). In C. virfiimca. Chintala and Fisher ( 1989) reported that oysters from an MSX-resistant line showed reduced suscep- tibility to MSX and higher serum lectin concentrations compared to native oysters, but were no less susceptible to P . nuiiiiuis than native animals. Only recently has attention been given to intraspecific variation in susceptibility to P. mannus. In a set of field trials in Chesa- peake Bay, Burreson (1991) reported that two strains of C. vir- ginica selected for increased resistance to MSX were highly sus- ceptible to P. marinus infection, failed to reach market size, and suffered 99% mortality over the course of the study. In contrast, three unselected (native I lines showed lower susceptibility to P. marinus and lower mortality (80%) and grew to market size. A second inbred strain of Chesapeake Bay origin, selected for growth rate but not disease resistance, also showed greater sus- ceptibility to P. marinus than native Chesapeake Bay or Delaware Bay oysters (Paynter and Burreson 1991). When cultured in North Carolina, this same selected strain showed faster growth than na- tive oysters but ceased growing and showed higher mortality than native oysters when P. marinus infection reached high levels. It is not clear whether the enhanced susceptibility to P. marinus in stocks selected for high growth rate or resistance to MSX repre- sents a genetic trade-off. or merely the negative consequences of inbreeding during the selection process, but there is clearly in- traspecific variation in response to P. marinus. Bushek (1994) found that oysters from different geographical regions show distinct differences in their response to infection w ith P. marinus (Figure 1). Following shell-cavity injections with P. marinus cultured in vitro, oysters originating from populations with previous exposure to the pathogen (Virginia and Texas) showed lower body burdens than oysters originating from less exposed populations (Maine and New Jersey). In addition, Atlan- tic Coast isolates of P . marinus caused heavier infection levels than Gulf Coast isolates. No interaction between virulence of the isolate and resistance of the oyster host population was observed, suggesting that the niechanism(s) of resistance to P. mannus may 100 80 60 40 20 ME NJ VA TX Oyster source Figure 1, Geometric mean infettiiin intensity Uells per g wet tissue) of oyster populations 94 cki\s after inoculation «ith in nVro— cultured P. marinus. Populations with a long histor> of exposure to P. marinus (Virginia and Texas! showed significantly IP = (1.004) lower infection intensity than populations with little or no history of exposure (Maine and New ,|ersey). From Bushek (1994). be general, rather than specific to particular strains of the patho- gen. La Peyre and Chu ( 1988) found that C. virf;inita from Mobjack Bay. VA (a site exposed to heavy MSX pressure), showed higher hcmocyte chemotaxis activity and agglutination titers than oysters from James River. VA (a source of oysters susceptible to both MSX and P. mannus). It is not known whether these differences reflect genetics, environmenlal influences, or both. One indirect way to examine the interaction between physiol- ogy and disease is to alter physiology by changing the number of chromosome sets an individual possesses (ploidy manipulation). Individuals with three sets of chromosomes (triploids) typically differ from their normal diploid counterparts by displaying re- duced gametogenesis and spawning. Do the physiological changes accompanying inhibited reproduction alter susceptibility to disease in Crassostreii'.' Triploidy did not appear to decrease susceptibility to P. mari- nus in either C. virginica or C. gigas exposed to the pathogen in flow-through seawater systems (Meyers et al. 1991 ). Field trials in Virginia also showed no effect of triploidy on susceptibility of C. virginica to P. marinus. but triploids did show higher growth rates (Barber and Mann 1991 ). Chu et al, (1993) reported that although C. gigas challenged with P. marinus trophozoites generally showed lower infection prevalence than C. virginica, triploid C. gigas held at 10°C showed a higher prevalence than either diploid C. gigas or C. virginica. At a commercial grow-out site in Massachusetts. Matthiessen and Davis (1991) found that triploid C. virginica showed higher rates of infection with MSX. but significantly lower mortality. than diploids. These differences may reflect the significantly larger size of the triploids rather than triploidy per se. In a large- scale field study, triploids exposed to both MSX and P. mannus 138 Gaffney and Bushek suffered consistently greater mortality than diploids, although the relative contributions of the two diseases have not yet been as- sessed (Allen and Gaffney, unpubl.). Interspecific Variation The limited data presently available suggest that differences between oyster species in resistance to pathogens such as P. nuiri- nus (and MSX) are far greater than intraspecific differences. Upon exposure to both MSX and P. marinus in a running seawater system in Maryland, native C . virginica died from both diseases, while C. gigas acquired no MSX infections and low levels of P. marinus, which did not intensify (Farley et al. 1991 ). In a similar study, Meyers et al. ( 1991 ) reported that 40% of C. gigas exposed to P. marinus became infected after 83 days, compared to 100% of C. virginica. Infections in C. gigas were light compared to those in C. virginica, and the low mortality observed in the former was attributed to causes other than P. marinus. Comparable results were obtained were obtained by Barber and Mann (1994). POPULATION STRUCTURE AND GENETIC VARIATION IN OYSTERS Subdivided Versus Panmictic Populations A panmictic population is one in which mating is random. Natural species may be subdivided into local subpopulations or denies, within which breeding may usually be treated as random. In the case of organisms with the capacity for extensive dispersal, such as oysters, it is difficult to determine the spatial boundaries of demes. Because it is virtually impossible to directly track the dispersal and fate of oyster larvae, we are focused to rely on indirect measures of population structure, i.e., genetic markers. Although C. virginica has been the subject of numerous genetic analyses of population structure, our understanding is far from complete. Patterns of mitochondrial DNA ImtDNA) variation sug- gest that C. virginica may be divided into distinct Atlantic Coast and Gulf Coast assemblages (Rceb and Avise 1990). In addition. allozyme surveys suggest that populations at the extremes of the species range (Nova Scotia and southern Texas) are genetically distinct from the main assemblages (see Gaffney 1996 for review). It is now becoming clear that these assemblages are further divided into subpopulations or "races." as many oyster biologists have long suspected. Barber et al. (1991) showed that at least some of the physiological differences among latitudinally sepa- rated Atlantic populations are genetic, i.e.. there are genetic dif- ferences in reproductive schedules between Long Island Sound oysters and Delaware Bay oysters. Preliminary data on mtDNA sequence variation likewise suggested that genetically distinct sub- populations may exist within the Atlantic and Gulf assemblages (O'Foighil et al., 1995). In a large-scale population survey of sequence variation in a 0.4-kb fragment of the small-subunit ( I6S) mitochondrial ribosomal DNA gene by denaturing gradient gel electrophoresis. Gaffney and Wakefield (unpubl.) found three ma- jor haplotypes, each of which was restricted to a single region (Gulf Coast, southern Atlantic, northern Atlantic). Within any region, haplotype diversity was negligible, with the exception of Prince Edward Island, which exhibited unusually high diversity. These results point to significant population subdivision, suggest- ing limited gene flow despite the potential for widespread larval dispersal. Further work is needed to characterize these genetically distinct oyster populations, particularly with regard to their re- sponse to disease. Population Structure and the Evolution of Disease Resistance From an evolutionary perspective, the interplay between host and pathogen is now realized to be more complex than originally thought and is the subject of continuing theoretical development. One important element is the population structure of the host spe- cies, i.e.. whether isolated subpopulation (demes) exist, and to what extent gene flow occurs among them. If gene tlow via larval dispersal is extensive, local adaptation may be retarded by the influx of genes from other areas subject to different selection pres- sures. This phenomenon is often invoked to explain the failure of C. virginica to develop effective resistance to pathogens such as MSX and P . marinus: despite locally intense disease pressure, the continued influx of immigrants from relatively uninfested source populations (and the export of locally produced offspring) limits the evolution of resistance in the host. The same argument was invoked to explain the failure of mainland populations of C . vir- ginica in Canada to develop increased resistance to the enigmatic Malpeque Bay Disease, despite introductions of resistant oysters from Prince Edward Island (Logic et al. 1960). DEVELOPMENT OF DISEASE-RESISTANT OYSTERS— PROBLEMS AND PROSPECTS We conclude, not surprisingly, that genetic variation for resis- tance to P . marinus is found within C. virginica, both within local populations and among geographic regions. An even greater de- gree of variation is found among Crassastrea species. It is possible to capitalize on this variability, to produce more disease-resistant oysters for aquaculture, or to replenish natural stocks? Developing resistant lines of C. virginica for aquaculture is relatively straightforward, at least in principle. Hatchery lines may be initiated with founders drawn from diverse sources, including areas with a long history of exposure to P. marinus. Selection may be imposed by culturing animals at known sites of heavy disease pressure. Appropriate breeding plans designed to increase disease resistance without compromising other production traits can be employed. Although years may be required to develop high- performance lines, it can be done with presently available tech- nology. Challenging oysters in the field with naturally occurring patho- gens is technically simple and is most likely to provide a realistic selection regime. This approach suffers, however, from unpredict- able interannual variation in disease pressure and the confounding effects of other variables (e.g.. other diseases, temperature and salinity variation). More controlled selection regimes can be em- ployed by deliberately infecting oysters in the laboratory. Ray ( 1954) demonstrated that uninfected oysters could be chal- lenged with P. marinus by sharing a tank with infected animals. This permits control of physical factors (e.g., temperature, salin- ity) but does not allow control of dose and exposure to other pathogens. This approach also provides little control over the ge- netic (clonal) composition of the pathogen population. Use of in v'(7ro-cultured P. marinus (La Peyre et al. 1993, Kleinschuster and Swink 1993, Gauthier and Vasta 1993) overcomes these problems but has other shortcomings. Bushek (1994) and Chintala et al. (1995) found in ivVro-cultured cells to be relatively uninfective when ingested by oysters (presumably the most common mode of infection). Injection either into the shell cavity or directly into tissue will produce infections (Gauthier and Vasta 1993. La Peyre et al. 1993, Bushek 1994). However, these methods bypass nat- ural barriers to infection (e.g., gill and palp sorting, tissue epithe- lia, etc.), which may contribute to differences in resistance among Genetic Aspects of Disease Resistance 139 oyster populations. At this point, field exposure to naturally oc- curring disease appears to be the best method for large-scale se- lection programs, while controlled laboratory infection will prob- ably prove valuable for mvestigating and more precisely evaluat- ing mechanisms of resistance in selected oysters. An alternative approach to traditional artificial selection in- volves attempting to move genetic elements that confer disease resistance from one species to another. For example. C. gigas appears to be highly resistant to both P. marunis and MSX. Can we move the gene(s) responsible from C . gigas to C. virginica? In principle, this is possible, even though we have no idea what genetic elements are responsible. A time-honored method is hy- bridization and subsequent backcrossing under selection for the desired trait. To date, this approach does not appear to be feasible, as hybrids between C. gigas and C. virginica are inviable (Allen et al. 1993). Modern genetic engineering techniques provide a means of moving genes between species that cannot be hybridized. This approach may be feasible once methods have been developed for producing transgenic bivalves and the gene(s) conferring disease resistance is identified. While the former barrier may not be dif- ficult to breach, the identification ot disease-resistance genes is not a trivial matter. A simpler form of genetic engineering, involving the transfer of chromosomes or chromosome fragments from C. gigas to C. virginica by means of partial gynogenesis, is currently under in- vestigation. When lightly irradiated C. gigas sperm are used to fertilize C. virginica eggs, the developing zygote will possess a single set of maternal chromosomes, plus any paternal chromo- somes or fragments that are not completely inactivated by irradi- ation. Application Of cytochalasin B to the newly fertilized egg blocks extrusion of the second polar body, resulting in a dip- loidized gynogenetic embryo — i.e.. one that contains two sets of maternally derived chromosomes, plus any paternal genes contrib- uted by the irradiated sperm. If the paternal fragments contain elements that provide disease resistance, and if these are incorpo- rated into the genome and stably expressed, the result may be a disease-resistant oyster. Efforts are currently underway to develop methods for partial gynogenesis in C. virginica (Guo. pers. comm.). None of these approaches to the development of disease- resistant C. virginica can be implemented quickly or simply. One alternative is to grow a disease-resistant Crassostrea species in areas now decimated by P. nninniis (or MSX). C. gigas has often been suggested as an alternative (Mann et al. 1991). in view of its worldwide use in aquaculture and resistance to both MSX and P. mariniis. Introduction of C. gigas has been credited with rescuing the cupped oyster industry in France, which crashed when the Portuguese oyster iCrassostrea angulata. considered either a con- specific or a close relative of C . gigas) succumbed to disease (Grizel and Heral 1991). However, its suitability for culture in mid- Atlantic waters is debatable, as field trials have shown high unexplained mortality (Barber and Mann 1994) and heavy shell damage from the boring polychaete Polydora (Burreson et al. 1994). In any event, it appears unlikely that C. gigas will be deliberately cultured in Atlantic waters, given the level of concern about introducing exotic species. A species of Crassostrea that exhibits good growth, viability, and disease resistance, yet fails to reproduce, is still a good can- didate for aquaculture in disease-ravaged areas. One possibility is the hybrid of C. gigas and Crassostrea rivularis. Unlike C. gigas, the latter species is apparently well adapted to high temperature and low salinity (Nie 1990). Viable hybrids can be produced (Allen and Gaffney 1993). If they prove to be sterile, and show good performance, they may someday earn a niche in commercial oyster culture. ACKNOWLEDGMENTS We thank Stan Allen for helpful comments on the manuscript. This work was supported in part by the National Marine Fisheries Service Oyster Disease Research Program. LITERATURE CITED Allen. S. K.. Jr. & P. M. Gaffney. 1493. Genetic confirmalion of hybrid- ization between Crassostrea gigas (Thunberg) and Crassostrea rivu- laris (Gould). Aquaculture 113:291-300. Allen. S. K.. Jr.. P. M. Gaffney. J. Scarpa & D. Bushek. 1993. Inviable hybrids of Crassostrea virginica (Gmelin) with C. rivularis (Gould) and C gigas (Thunberg). Aquaculture 113:269-289. Alvarez. G.. C. Zapata. R. Amaro & A. Guerra. 1989. Multilocus het- erozygosity at protein loci and fitness in the European oyster. Ostrea edulis L. Heredity 63:359-372. Barber. B. J.. S. E. Ford & D. T. J. Littlewood, 1991a. A physiological comparison of resistant and susceptible oysters Crassostrea virf^inica (GmelinI exposed to the endoparasite Haplosporidium nelsoni (Haskin. Stauber & Mackin). J. Exp. Mar. Biol. Ecol. 146:101-112. Barber. B. J.. S. E. Ford & R. N. Wargo. 1991b. Genetic variation in the timing of gonadal maturation and spawning of the eastern oyster. Cras- sostrea virginica (Gmelin). Biol. Bull. 181:216-221. Barber. B. J. & R. Mann. 1991. Sterile triploid Crassostrea virginica (Gmelin. 1791 ) grow faster than diploids but are equally susceptible to Perkinsus marinus. J. Shellfish Res. IO:443-;50. Barber. B. J. & R. Mann. 1994. Growth and mortality of eastern oysters. Crassostrea virginica (Gmelin. 1791). and Pacific oysters. Crassos- trea gigas (Thunberg. 1793) under challenge from the parasite. Per- kinsus marinus. J. Shellfish Res. 13:109-1 14. Blol. M. & C. Thiriot-Quievreux. 1989. Multiple locus fitness in a transfer of adult Mvtilus desolationis (Mollusca. Bivalvia). In: J, S. Rvland & P. A. Tyler (eds). Reproduction. Genetics and Distributions of Ma- rine Organisms. Olsen & Olsen. Fredensborg. Denmark, pp. 259-264. Borsa. P.. Y. Jousselin & B. Delay. 1992. Relationships between allo- zymic heterozygosity, body size, and survival to natural anoxic stress in the palourde Ruditapes decussatus L. (Bivalvia: Veneridae). J. Exp. Mar. Biol. Ecol. 155:169-181. Bunell. V. G.. Jr.. M. Y. Bobo & J. J. Manzi. 1984. A comparison of seasonal incidence and intensity oi Perkinsus marinus between subtidal and interidal oyster populations m South Carolina. J World Maricul Soc. 15:301-309. Burreson. E. M. 1991. Effects oi Perkinsus marinus infection in the east- em oyster. Crassostrea virginica (Gmelin. 1791): I. Susceptibility of native and MSX-resistant stocks. J. Shellfish Res. 10:417^23. Burreson. E. M.. R. Mann & S. K. Allen. Jr. 1994. Field exposure of triploid Crassostrea gigas to Haplosporidium nelsoni (MS.X) and Per- kinsus marinus (Dermo) in the lower Chesapeake Bay. J. Shellfish Res. 13:293. Bushek, D. 1994. Dermo disease in American oysters: genetics of host- parasite interactions. Ph. D. Dissertation. Rutgers. New Brunswick. New Jersey. 189 pp. Chintala. M. M.. K. A. Alcox. S. E. Ford & D. Bushek. 1995. Cultured and natural Perkinsus marinus cells: a possible mechanism for viru- lence differences. 15th Milford Aquaculture Seminar. Feb. 21-23. 1995. Milford. Connecticut (abstract). 140 Gaffney and Bushek Chintala. M. M. & W. S. Fisher. 1989. Comparison of oyster defense mechanisms for MSX-resistant and -susceptible stocks held in Chesa- peake Bay. J. Shellfish Res. 8:467-468. Chu, F. E.. C. E. Burreson, A. Volety & G. Constantin. 1993. Perkmsus marinus susceptibility in eastern (Crassostrea virginica) and Pacific (Crassoslrea gigas) oysters: temperature and salinity effects. J Shell- fish Res. 12:360 (abstract). Chu. F. E. & J. F. La Peyre. 1993. Development of disease caused by the parasite. Perkinsus marinus and defense-related hemolymph factors m three populations of oysters from the Chesapeake Bay. USA. J Shell- fish Res. 12:21-27. Ewart. J. W.. R. Cole & J. Tinsman. 1988. Growth and survival of hatch- ery produced MSX resistant oysters stocks in Delaware Bay. J. Shell- fish Res. 7:543. Farley. C. A.. D. L. Plutschak & G. E. Krantz. 1991 . Crassoslrea gigas disease exposure to Haplosporidiwn nelsoni and Perkinsus marinus in Chesapeake Bay waters. J. Shellfish Res. 10:306-307, Fevolden. S. E. 1992. Allozymic variability in the Iceland scallop Chlamys islandica: geographic variation and lack of growth- heterozygosity correlations. Mar. Ecol. Prog. Ser. 85:259-268. Fevolden. S. E.. T. Refstie & K. H. Roed. 1992. Disease resistance in rainbow trout {Oncorhyncus mykiss) selected for stress response. Aqua- culture 104:19-29. Ford, S. E. & A. J. Figueras. 1988. Effects of sublethal infection by the parasite Haplosporidium nelsoni (MSX) on gametogenesis. spawning, and se.x ratios of oysters in Delaware Bay. USA. Dis. Aquai. Org. 4:121-133. Ford. S. E.. A. J. Figueras & H. H. Haskin. 1990, Influence of selective breeding, geographic origin, and disease on gametogenesis and sex ratios of oysters. Crassoslrea virginica. exposed to the parasite Hap- losporidium nelsoni MSX. Aquacullure 88:285-301. Gaffney. P. M. 1986. Physiological genetics of growth in marine bivalves. Ph.D. Dissertation. State University of New York at Stony Brook. Stony Brook. New York. 305 pp. Gaffney. P. M. 1990. Enzyme heterozygosity, growth rate, and viability in Mylilus edulis: another look. Evolulion 44:204-210. Gaffney. P. M. 1996. Biochemical and population genetics. In: The East- em Oyster: Crassoslrea virginica. Maryland Sea Grant College Pro- gram, in press. Gauthier, J. D. & G. R. Vasta. 1993. Continuous in vilro culture of the eastern oyster parasite Perkinsus marinus. J. Inverlehr. Palhol. 62: 321-323. Grizel. H. & M. Heral. 1991. Introduction into France of the Japanese oyster (Crassoslrea gigas). J. Cons. Int. Explor. Mer. 47:399-403. Haskin. H. H. & S. E. Ford. 1979. Development of resistance to Min- chinia nelsoni (MSX) mortality in laboratory -reared and native oyster stocks in Delaware Bay. Mar. Fish. Rev. 41:54-63. Haskin. H. H. & S. E.Ford. 1988. Characteristics of inbred oyster strains selected for resistance to Haplosporidium nelsoni (MSX). J. Shellfish Res 7:162. Hawes. R. O.. K. Scully & H. Hidu, 1990. Growth rate of two diverse populations of American oysters. Crassoslrea virginica, and their re- ciprocal crosses. Aquacullure 85:327. Hawkins. A, J. S.. B. L. Bayne. A. J. Day. J. Rusin & C. M. Worrall. 1989. Genotype-dependent interrelations between energy metabolism, protein metabolism and fitness. In: J. S. Ryland & P. A. Tyler (eds). Reproduclion. Genelics and Dislribulions of Marine Organisms. Olsen & Olsen. Fredensborg. Denmark, pp. 283-292. Hopkins, S. H. 1954. Oyster setting on the Gulf Coast. Proc. Nail. Shell- fish. Assoc. 45:52-55. Kleinschuster. S. J. & S. L. Swink. 1993. A simple method for the in vitro culture of Perkinsus marinus. Nautilus 107:76-78. Koehn, R. K. & S. E, Shumway. 1982. A genetic/physiological explana- tion for differential growth rate among individuals of the American oyster. Crassostrea virginica (Gmelin). Mar. Biol. Lett. 3:35—42. La Peyre. J. F. & F. E. Chu. 1988. Hemocyte and humoral activities of two populations of oyster (Crassostrea virginica). J. Shellfish Res. 7:544. La Peyre. J. F.. M. Faisal & E. M. Burreson. 1993. In vitro propagation of the protozoan parasite Perkinsus marinus. a pathogen of the eastern oyster. Crassostrea virginica. J. Euk. Microbiol. 40:304-310. Logic, R. R., R. E. Drinnan & E. B. Henderson. I960. Rehabilitation of disease-depleted oyster populations in eastern Canada. Proc. Gulf Car- ibb. Fish. Inst. 13:109-113, Mann. R.. E. M. Burreson & P. K. Baker. 1991. The dechne of the Virginia oyster fishery in Chesapeake Bay: considerations for intro- duction of a non-endemic species. Crassoslrea gigas (Thunberg, 1793). J. Shellfish Res. 10:379-388. Matthiessen, G, C. & J. P. Davis. 1991. Growth, disease resistance and survivorship of diploid and triploid eastern oysters in an MSX envi- ronment. J. Shellfish Res. 10:298-299, Matthiessen. G. C, S. Y. Feng & L. Leibovilz. 1990. Patterns of MSX (Haplosporidium nelsoni infection and subsequent mortality in resis- tant and susceptible strains of the eastern oyster. Crassoslrea virginica (Gmelin. 1791). in New England. J. Shellfish Res. 9:359-365. Meyers. J. A.. E. M. Burreson. B. J. Barber & R. Mann. 1991. Suscep- tibility of diploid and triploid Pacific oysters. Crassoslrea gigas (Thun- berg. 1793) and eastern oysters. Crassoslrea virginica (Gmelin. 1791 ). to Perkinsus marinus. J. Shellfish Res. 10:433—437. Nie. Z.-Q. 1990. The culture of manne bivalve mollusks in China. In: W. Menzel (ed.). Estuarine and Marine Bivalve Mollusk Culture. CRC Press. Boca Raton, Florida, pp. 261-276. OFoighil. D.. P. M. Gaffney & T. J. Hilbish. 1995. Differences in mitochondrial 16S ribosomal gene sequences allow discrimination among American (Crassoslrea virginica Gmelin) and Asian (C. gigas Thunberg. C. ariakensis Wakiya) oyster species. J. E.xp. Mar. Biol. Ecol. 192:211-220. Paynler. K. T, & E, M, Burreson, 1991. Effects of Perkinsus marinus infection in the eastern oyster. Crassoslrea virginica: II. Disease de- velopment and impact on growth rate at different salinities. J Shellfish Res. 10:425-431 Paynter, K. T. & L. DiMichele. 1990. Growth of tray cultured oysters (Crassostrea virginica Gmelin) in Chesapeake Bay. Aquacullure 87: 289-298. Pecon Slattery. J., R. A Lutz ii R. C, Vnjenhoek. 1993. Repeatability of correlations between heterozygosity, growth and survival in a natural population of the hard clam Mercenaria mercenaria L. J . E.xp. Mar. Biol. Ecol. 165:209-224. Ray. S. M. 1954. Biological studies of Dermocystidium marinum. a fungal parasite of oysters. Rice Institute Pamphlet. Special Issue 1 14 pp. Reeb. C. A. & J. C. Avise. 1990. A genetic discontinuity in a continu- ously distributed species: mitochondrial DNA in the American oyster. Crassoslrea virginica. Genelics 124:397^06. Rodhouse. P. G. & P. M. Gaffney. 1984. Effect of heterozygosity on metabolism during starvation in the American oyster Crassostrea vir- ginica. Mar. Biol. 80:179-187. Rodhouse. P. G.. J. H. McDonald. R. I. E. Newell & R. K. Koehn. 1986. Gamete production, somatic growth and multiple-locus het- erozygosity in Mylilus edulis. Mar. Biol. 90:209-214. Van Valen. L. 1973. A new evolutionary law. Evol. Theory 1:1-30. Widdows. J.. P. Donkin. P N Salkeld. J. J. Cleary. D. M. Lowe. S. V Evans & P. E. Thomson. 1984. Relative importance of environmen- tal factors in determining physiological differences between two pop- ulations of mussels (Mylilus edulis). Mar. Ecol. Prog. Ser. 17:33- 47. Wilson, E. A.. E. N. Powell. M. A. Craig. T. L. Wade & J. M. Brooks. 1990. The distribution o( Perkinsus marinus in Gulf Coast oysters: its relationship with temperature, reproduction, and pollutant body bur- den. Int. Rev. Gesamien Hvdrobiol. 75:535-550. Zouros, E. & D. W. Foltz. 1987. The use of allelic isozyme variation for the study of heterosis. Isozymes Curr. Top. Biol. Med. Res. 13:1- 59. Zouros, E., S. M. Singh. D. W. Foltz & A. L. Mallet. 1983. Post- settlement viability in the American oyster (Crassostrea virginica): and overdominant phenotype. Genet. Res. 41:259-270. Journal of Shellfish Research. Vol. 15, No. 1. 141-165. 1996. MODELING DISEASED OYSTER POPULATIONS. II. TRIGGERING MECHANISMS FOR PERKINSUS MARINUS EPIZOOTICS ERIC N. POWELL,' ' JOHN M. KLINCK,^ AND EILEEN E. HOFMANN- ^ Department of Oceanography Te.xas A&M University College Station, Texas 77843 'Center for Coastal Physical Oceanography Crittenton Hall Old Dominion University Norfolk. Virginia 23529 ABSTR.ACT Densities of Crassoslrca virginica remain high enough to support substantial fisheries throughout the Gulf of Mexico despite high mortality rates produced by the endoparasite Perkinsiis marinus. The infrequency of epizootics in these populations suggests that controls exist on the disease intensification process. The progression of epizootics in oyster populations, the factors that trigger epizootics, and the factors that terminate epizootics once started were investigated with a coupled oyster population-P. marinus model. The time development of a simulated epizootic was triggered by environmental conditions that occurred and disappeared as much as 18 months prior to the onset of mortality in the oyster population. Initiation of epizootic conditions was detected as an increase in infection intensity in the submarket-size adult and juvenile portions of the oyster population. Infection intensity of the market-size adults IS maintained at a relatively stable level by the death of heavily infected individuals and the slow rate of P. marinus division at high infection intensities. Once started, most of the simulated epizootics resulted in population extinction in 2 to 4 years Stopping an epizootic required reducing the infection intensity in the submarket-size adults and juveniles. The infection intensity of market-size adults does not need to be reduced to stop an epizootic nor must it be raised to start one The simulated oyster populations show that a reduction in ingestion rate (by reduced food supply or increased turbidity) can trigger an epizootic, especially if the reduction occurs during the summer. Increasing food supply or decreasing turbidity in the following year does not necessarily prevent the occurrence of an epizootic. Rather, the onset of the event is simply delayed. Additional simulations show that the relative combination of vanations in salinity and temperature is important in determining the occurrence of an epizootic. A dry (high-salinity) summer followed by a warm winter produces conditions that favor the development of an epizootic. Conversely, a warm dry year followed by a cool wet year fails to produce an epizootic. Simulations that consider variations in the biological charactenstics of oyster populations, such as changes in recruitment rate or disease resistance, show that these are important in regulating the occurrence of an epizootic as well as in terminating the event. In particular, increased recruitment rate dilutes the infected population sufficiently to terminate an epizootic. One primary conclusion that can be obtained from these simulations is that epizootics of P. marinus in oyster populations are difficult to generate simply with changes in either temperature or salinity. Rather, the epizootics are triggered by some other factor, such as reduced food supply or reduced recruitment rate, that occurs prior to or coincident with high salinity or temperature conditions. KEY WORDS: Perkinsus nuiriniis disease, disease model, oyster disease, eastern oysters. Crassoslrea virginica INTRODUCTION fisheries throughout much of the range of this animal (Hofstetter 1990. NOAA 1991. Powell et ai. 1995b). Throughout the southern extent of their habitat range, popula- Nevertheless, epizootics produced by P. marinus do occasion- tions of the eastern ovster, Cra.s.sostrea virgimca. arc impacted ally occur throughout the range of P. marinus. although those greatly by the disease-producing endoparasite Perkinsu.s marnws occurring in the mid-Atlantic region have been more noteworthy in (Quick and Mackin 1971. Wilson et al. 1990. Lewis et ai. 1992). 'heir areal extent and effect on the fishery (Mann et al. 1991. In the Gulf of Mexico, the market-size component of oyster pop- Sindermann 1993). Epizootics of most animal species follow a ulations generally suffers about 50% yearly mortality due to P. series of characteristic stages (Gill 1928). which are shown sche- marinus m-dckm 1961, 1962, Hofstetter 19771. Only one Gulf of matically in Figure 1. Most are triggered during a preepizootic Mexico oyster population is known to be free of infection from P. phase when the host population appears to be at its healthiest marinus (Powell et al. 1992a). Typically, prevalence of this or- -' f egg Larval recruitment Larvae mortality Number of larvae recruited spawn ' = .s (number of eggs spawned) Postsettlement population natural mortality M^ = ipfnumber of living), for; = k.l Postsettlement salinity mortality M, = A'; (number of living) i, = fa, S + p|)r + (a,S + pj Caloric conversions Oysters Food Oyster eggs at 7 s= 20°C atS S3 15 ppt at 10 ppt < 5 < 15 ppt at 5 « 10 ppt 0.39 g ash-free dry weight, about 50 mm Reproductive tissue development for a given oyster size class as a function of reproductive efficiency, R^t^. and total net production. NP,, Reproductive efficiency temperature dependence for January to June Reproductive efficiency temperature dependence for July to December Preferential resorption of gonadal tissue Spawning occurs when the reproductive biomass exceeds 20% of total oyster biomass f,^„„. the ratio of females to males; L,,. length in mm Number of eggs spawned, C is number of calories per egg, W^^^ is egg weight V^gg, oyster egg volume Larval planktonic time assumed to be 20 days s. the mortahty rate, in spawn ' Mp. the number dying time ' k^,. the daily mortality rate ((/ '); k and /. the inclusive size classes being affected by mortality M,. the number dying time ' A',, daily mortality rate id ') a, = -0.000348 a, = 0.00232 Pi = 0.01764 P, = -0.3089 5. ambient salinity (ppt) 7. ambient temperature ("C) 6100 cal (g dry wt) ' 5168 cal (g dry wt) ' 6133 cal (g dry wt) ' observed by Saunders et al. (1993) ranged between 4 and 10 hours at30°C and 17 ppt. Given the limited observations on P. maiiinis growtii in vivo, this process was modeled using standard relationships for temperature and salinity dependencies which were calibrated by comparing the sim- ulated growth of P. marinus to data sets that provide observations of the seasonal dependency of parasite infection intensity as salinity and temperature change. These data came from April Fools Reef in Galveston Bay, TX (Soniat 1985), Biloxi Bay, MS (Ogle and Flurry 1980), and North Inlet, SC (Crosby and Roberts 1990). Temperature control on the specific rate of parasite division, /;/r), was assumed to follow a standard exponential form: TABLE 3. CoefTicient dennitions and values for the mussel model. Coefficient Definition Value Units a h Mussel filtration rate Mussel weight Mussel length Mussel weight scaling factor Mussel weight scaling factor Calculated ml mussel ' min ' Calculated g dry wt Assigned mm -4.8979 No units 2-8734 No units 148 Powell et al. rj(T) = r,,,?"'^'"-^"'. (11) To calibrate equation (11), a known division rate at a given tem- perature is needed. Observations of field populations suggest that infection intensity begins to rise in most populations when the temperature exceeds 20°C and the salinity exceeds 20 ppt. There- fore, the 20°C-20 ppt boundary was used to standardize parasite division and mortality rates. At 20°C and 20 ppt, parasite division should just balance loss (Ray 1954. Mackin 1962, Andrews 1988). The division time at 20°C and 20 ppt was set at 30 hours by comparing simulated distributions to those in Soniat ( 1985), Ogle and Flurry (1980), and Crosby and Roberts (1990). This division time is within the ranges of those reported from the limited labo- ratory and in vivo measurements. A (?,,, of 2.0. which is consistent with measurements for P, inwimis (Chu and Greene 1 989 1, is used to calculate a parasite division rate at temperatures other than 20°C. The coefficients and their values thus determined for equa- tion ( 1 1 ) are defined in Table 4. The rate of parasite division is independent of salinity except at and below 10 ppt (Chu and Greene 1989. Ragone and Burreson 1993). Thus, for salinities (5) below 10 ppt. equation (II) is modified as: rjiT.S) = r,n 10 ,amt)~Ta) (12) where coefficient definitions and values are given in Table 4. This relationship provides a decrease in parasite division rate at low salinity but retains the temperature relationship. Simulations of P . marinus population dynamics using equation (12) in the oyster-P. marinus model resulted in parasite growth rates and densities that were too high relative to those suggested by field measurements in Soniat ( 1985). Ogle and Flurry (1980). and Crosby and Roberts (1990) under the appropriate environmental constraints (Hofmann et al. 1995). Most measurements of proto- zoa in culture show that parasite division rate decreases at high population densities as food becomes limiting (Hall 1967). A sim- ilar response by P. marinus is suggested by in vivo experiments in which the rate of DNA production by P. marinus at various par- asite densities declined at high densities (Saunders et al. 1993). Also, a decrease in hemolymph protein in oysters has been noted during summer months when P. marinus infection intensity is high (Chintala and Fisher 1991 ) and as a result of MSX infection (Ford 1986). Usinc the measurements from Saunders et al. (1993), an TABLE 4. Coefficient definitions and values for the P. marinus population model. Coefficient Definition Value Units r/T) rju a To So P 1 rjT.S) ''ma S Ec Eg Er El e D i e K \ M- V z 1 a V T r, r,b Specific rate of parasite division Base specific parasite division rate (?,„ conversion Base temperature for parasite division rate Base salinity for parasite division rate Base specific parasite division rate Parasite density scaling factor Parasite number Oyster weight Parasite density scaling factor Specific parasite loss rate Base specific parasite loss rate (2i(, conversion Total P. marinus energy demand Energy for P. marinus population increase Energy forf. marinus respiration demand P. marinus mortality Conversion Average parasite cell diameter Conversion Respiration scaling factor Respiration scaling factor Filtration scaling factor Filtration scaling factor Conversion Filtration rate, infected oyster Lethal parasite density Mortality scaling factor Weight scaling factor Weight scaling factor Mortality scaling factor Weight conversion factor Infection level scaling factor Infection level scaling factor Specific rate of transmission Base specific interpopulation transmission rate Base specific intrapopulation transmisson rate Calculated 0.555 0.06931 20 20 0.555 2.454 X 10* Calculated Table 1 -1.5 Calculated 0.555 0.08153 Calculated Calculated Calculated Calculated 1.16 X 10^ 8 9.57 X 10"' -4.09 0.75 0.58 579 -2.287 X Calculated Calculated 2.057 1.3258 X 0.2625 3.2 S 1409.9 0.64296 Calculated 0.2 12 10" 10" d' d-' 0(-.-l °C ppt d ' g AFDW cell " ' Number of cells g AFDW No units d ' d' T-' cal d " ' cald"' cald"' cal d - ' hr cal d " ' nr ' |j.m cal |j.m"^ ml hr" ' |xm"^ No units No units No units g AFDW ceir' ml oyster ' min~ ' Cells oyster" ' No units g AFDW No units No units g wet wt (g dry wt)" Cells (g wet wt)" ' No units d-' y-' y-' Triggering of Pf.rkinsus marinus Epizootics 149 empirical relationship that modifies the specific parasite division rate at high parasite density, ''./(p), j, was derived as: rd(p),.k = ^rja.S) (13) where r,(r,5) is determined from equation (12). Coefficient def- initions and values are given in Table 4. In the model, the parasite division rate that is used is the minimum of that determined from equations (12) and (13). Perkinsus marinus Mortality Mortality of P. marinus is presumably a result of the oyster defense system response. Thus, parasite mortality was parameter- ized using data obtained for hemocytes. which are an important component of the oyster's defense mechanism (Fisher 1S)S8). These data show that parasite mortality is temperature and salinity dependent. Moreover, field (Soniat 1985. Burrell et al. 1984) and laboratory (Fisher et al. 1992) observations show that the effect of salinity on parasite mortality is discernible only at high tempera- tures. One explanation for this is that hemocytes are already max- imally active at low temperature so that salinity changes have little effect. However, at higher temperatures where hemocyte activity is reduced, some capability is recovered when the oyster is ex- posed to low salinity. Thus, a temperature- and salinity-dependent relationship for the specific parasite mortality rate. rjT.S). was obtained using measurements of hemocyte activity reported in Fisher and Newell ( 1986). Fisher and Tamplin (1988). Fisher et al. ( 1989. 1992). and Chintala and Fisher (1991 ) as: rjT.S) =. r„,o e -(f') /r„-iOi -5 —J — (Sin-S.i) (14) where e is the larger of the temperatures at 10°C or the difference between the ambient temperature and the base temperature of 20°C. i.e.. max(l()°C,r(n - T„). The definitions and values for the coefficients in the above equation are given in Table 4. The value used for & was obtained by applying a gio of 2.26 to the base mortality rate. The specific parasite mortality rate assumes no reduction in hemocyte activity at extreme low salinity. Ford and Haskin (1988) found active hemocytes down to 6 ppt. and oyster mortality from low salinity begins at lower salinities. As with parasite division, mortality should also be dependent on parasite density. As parasite density increases, the effectiveness of the defense system should decrease. Anderson et al. (1992) showed that the number of hemocytes in heavily infected oysters is only about double that in lightly infected oysters, whereas the number of P. marinus cells is a factor of 1000 or more higher. Thus, the relative activity of the hemocytes must decline at high parasite density. Measurements sufficient to exactly describe the relationship between parasite concentration and mortality are not available. Hence, the parasite density effect on mortality rate was assumed to follow the same relationship as was used for the par- asite density effect (equation 13) on parasite division rate: ^„(p)y.^ = |3r„,(r,5)(^ (15) resistant to environmental extremes as its oyster host (Goggin et al. 1990) and calibrating simulations off. marinus growth against existing data sets did not require an additional mortality source beyond that provided by the host's defense system (Hofmann et al. 1995). Perkinsus marinus Energy Demand The P. marinus population depends on the oyster host to pro- vide sufficient energy to support parasite respiration and growth. Thus, the energy requirement of the parasite population, Ec. can be expressed as: Coefficient definitions and values are given in Table 4. The spe- cific parasite mortality rate was taken to be the minimum of the rate calculated using equations (14) and (15). It is assumed that no P. marinus mortality occurs as a direct result of extremes in tem- perature and salinity. Available data suggest that P. marinus is as Ec Eg + Er El (16) where Ei; is the energy required to increase the population biomass through parasite division and Er is the energy requirement for population respiration. The last term on the right of equation (16), El, represents the return of energy to the host from the parasite which occurs through parasite mortality. Although hemocyte exomigration (Cheng 1983) might limit the importance of El. exomigration is not included in the model. The terms in the above equation are formulated as described below. The energy requirement for population growth is defined by Eg - El and is determined by the net change in parasite number (C, j.) in the P. marinus population in a specific time interval. This is calculated from the difference in the specific parasite division and mortality rates as: ''Cj.k rjp),k) C,,k- (17) The change in parasite number in a time interval. AC, j., obtained from the above equation is converted to calories exchanged be- tween the parasite and its host by: £?,.. - El,, (18) e V AC,,, where e is a conversion factor obtained by assuming that 5 g wet weight is equivalent to 1 g dry weight and that 20 joules is equiv- alent to 1 mg dry weight (Layboum-Parry 1987). Parasite cell volume. V. is calculated as: 5" D (19) The average cell diameter. D. is from Ray (1954). Coefficient values and definitions are given in Table 4. The respiratory energy required by the P marinus population is obtained from: Er,,, = i e iamn-T)„) io-v«c,. (20) where the conversion factor, l,. assumes 4.83 ml O, per calorie (Powell and Stanton 1985). The exponents w and 6, which scale respiration rate to parasite cell volume, are from measurements made for protozoa (Fenchcl and Finlay 1983). The value for a assumes a ^k. oi 2 (Layboum-Parry 1987). The effect of salinity on P. marinus respiration rate is unknown and therefore is not included. Coefficient values and definitions are given in Table 4. Effects of Perkinsus marinus on Oyster Physiology The primary effects of P. marinus infection on oysters are to reduce oyster filtration rate (Lund 1957) and eventually cause host mortality. Although increased predation is frequently described as a product of parasitism (Jakobsen et al. 1988. Hadeler and Freed- 150 Powell et al. man 1989, Schmid-Hcmpel and Schniid-Hempel 1988). no evi- dence exists for selective predation of P. HKin/ii/.v-infected indi- viduals. Thus, selective predation is not included in the model. Also, the possible loss of P. marinus during spawning (Dungan and Roberson 1993) is not included. Mackin and Ray (1955) provide measurements of P. marinus that can be used to derive a relationship that describes the reduc- tion in oyster filtration rate with infection intensity. These mea- surements show an exponential decrease in oyster filtration rate that depends on the ratio of the number of cells of the parasite to the size (weight) of the host. This reduction in filtration can be expressed as: (21) Q = io'"'<'S'"l (24) Dredjk = o Ke^w, + 1 Coefficient definitions and values are given in Table 4. The ex- pression given in equation (21 ). when applied to the oyster filtra- tion rate, FR^ . defined in Table 2, results in a fractional reduction in filtration rate as: FRr FR,i\ - Dred,^) (22) where FR,, ^ is the filtration rate that results when the oysters are infected with P. marinus. The level of P. marinus infection in an oyster population is typically diagnosed in terms of a 0- to 5-point scale that was developed by Mackin (1962), with 5 being the heaviest infection level. Field and laboratory measurements show that oyster mor- tality generally occurs in individuals that have an infection inten- sity that corresponds to a 5 on this scale (Andrews 1988). Popu- lations with mean infection intensities of 3 or more generally suf- fer 50 to 75% mortality per year (Ray and Chandler 1955, Mackin 1961, Mackin and Hopkins 1961). These observations provide a basis for determining the lethal P. marinus infection level in the simulated oyster populations. A relationship was developed between host mortality, host size, and P. marinus number by assuming that host mortality occurs when the energy demand of the P. marinus population is some fraction of the host's net production. This relationship is based on net production values calculated for uninfected oysters as described by White et al. (1988) and is of the form: Ec (23) where NP,, is net production. Ec is the caloric requirement of the P. marinus population as determined from equation ( 14), and C^ is the lethal parasite density (cells oyster"') for any oyster size class, j. The above equation allows for a size dependency in lethal parasite density that is suggested by measurements given in Choi et al. (1989) and is consistent with a size-dependent scope for growth in oyster populations (Hofmann et al. 1992). The factor of 2 used in equation (21) was determined empirically by using yearly mortality rates of 90. 50. and 10% for the market-size population and comparing the resulting simulated populations with oyster populations reported in the field studies by Ogle and Flurry (1980), Soniat (1985), and Crosby and Roberts (1990). The lethal parasite density from equation (23) can then be related to oyster size through a regression of the form; Note that because equation (24) is obtained from a regression, the units on the two sides of the equation are not equivalent. Coeffi- cient definitions and values are given in Table 4. Assuming that P. marinus infections are initiated by one cell, then depending on oyster size, 22 to 27 population doublings are needed to reach the lethal density. Smaller oysters require fewer population doublings to reach the lethal parasite level. As required by field observations, the above equation yields a value of 5 on Mackin's Scale when converted according to Choi et al. ( 1989) as: C, = 1/7(10*")^, (25) where M is the Mackin's Scale infection intensity as defined by Craig et al. ( 1989). Coefficient values and definitions are given in Table 4. Equation (24) is consistent with the suggestion that oyster mor- tality could be at least partly explained by a negative energy bud- get produced when the energy demand of P. marinus exceeds the assimilation rate of the oyster (Choi et al. 1989). However, the exact mechanism by which P. marinus causes mortality of the oyster host is unknown, and some studies have reported significant effects on the host at lower infection levels (e.g., Paynter and Burreson 1991). The justification for using the approach given above comes from favorable comparisons between simulated and observed levels of P. marinus infection under equivalent environ- mental conditions (Hofmann et al. 1995). Perkinsus marinus Transmission The available studies of the transmission of P. marnnis indicate that oyster density and distance between infected host populations affect the rate of infection (Andrews and Ray 1988, Ford 1992, Mackin 1952). However, little information on the transmission of this disease from controlled experiments is available (Andrews 1965. 1988). Therefore, the transmission of P. marinus was mod- eled using general relationships for disease transmission. These formulations were then calibrated against field data. The specific rate of infection of uninfected oyster individuals. r,, was assumed to be the result of an interpopulation transmission rate, r,^. and an intrapopulation specific transmission rate, r^, as: '■( ^ 0(1 + '■(•1 (26) where P,, P,, and P, are factors that modify the intrapopulation transmission rate. Insufficient data were available to include the expected rela- tionship between oyster filtration rate and P. marinus transmission rate as occurs in other host-parasite systems (e.g.. Gee and Davey, 1986). This effect could be important at higher latitudes where filtration ceases during the winter, thus limiting transmission rate. However, the decrease in P. marinus prevalence and infection intensity produced by the effects of low temperature on parasite growth and mortality, that occur during the winter, should mini- mize any error due to exclusion of this effect. Also, a suspected intluence of salinity on P. marinus transmission rale (Paynter and Burreson 1991, Chu and La Peyre 1993) is not included in the model. The interpopulation infection intensity was determined by us- ing observations from San Antonio Bay, TX, obtained as part of the NOAA National Status and Trends program. A catastrophic Triggering of Phrkinsus marinus Epizootics flood produced lOO'/f mortality of oysters in this bay in 1988. As the bay recovered, the infection intensity and prevalence of P. marinus were monitored in the oyster population. These observa- tions showed that P. marinus infection returned to regional norms in about 2 years. Simulations of this event required an interpop- ulation infection intensity {r,i,) of 0.2 y"'. Field experiments by Paynter and Burreson (1991) yielded similar results. Three variables — oyster density. P . marinus prevalence, and P . marinus infection intensity — were used to determine the intra- population transmission rate. Factors affecting the intrapopulation transmission rate were formulated as follows. The prevalence of infection in a population varies between 0 and I . where 0 repre- sents an uninfected population and I represents a population in which all individuals are infected. At each time step m the model the fraction of the total population that was infected with P . mari- nus was calculated and this value was used to specify P^ as: fraction infected (27) Mean population infection intensities of 3.5 and above on Mackin's Scale are associated with substantial oyster mortality. Mortality should ma.ximize transmission rate by releasing infective elements into the water column where they arc transmitted to other individuals. Thus. P, was specified by establishing a ratio between the total parasite density. TCD, in the simulated oyster population and the parasite density that corresponds to an infection level (IL) of 3.5. Limiting the maximum value of this ratio to 1 yields a maximum transmission rate at all population infection intensities 2=3.5 of: TCD IL where TCD = ■^A=l *-< -j=i - Ojv (28) (29) and IL corresponds to 2.5 x 10^ cells (g wet wt)" '. The proximity of oyster individuals to one another also affects the rate of disease transmission. This effect is included by com- paring the total simulated oyster population density with a rela- tively high oyster population density and limiting this value to a maximum of I as: P^ = min 1.. II 28 ;=1 k=\ OD (30) where OD is 4000 oysters m'- (May 1971. Dame 1976). A dense, heavily infected population [(P, + Pi + P}^^ = '1 should produce an intrapopulation transmission rate that is capable of infecting all uninfected individuals within 6 months. To achieve this effect, the maximum intrapopulation transmission rate (/„,) was set to 1 2 y " ' . Model Implementation The oyster and P. marinus model described above requires input of environmental measurements that describe ambient food supply, turbidity level, current flow velocity, salinity, and tem- perature conditions. For this study, time series of these data were constructed to illustrate specific environmental effects. In all cases, the structure of the environmental time series was based on measurements made in Galveston Bay. TX. The time series con- sist of monthly averaged values that extended for 1 year. The various environmental time series are given in Table 5. The spe- cific combination of the environmental series for the different simulations is given in Table 6. The oyster population-P. marinus model was solved numeri- cally using an implicit (Crank-Nicolson) tridiagonal solution tech- nique with a 1-day time step. All simulations began on January I (Julian day 1) and ran for 6 years. This amount of time was sufficient for the oyster and parasite population to adjust to the environmental forcing. Each simulation was initialized with an oyster size-frequency distribution obtained from a reef. South Deer Island, in the West Bay section of Galveston Bay. TX, in spring 1992 (Fig. 3). The initial density of the individuals in the oyster population was set at 20 individuals m " ~. The mussel size- frequency distribution used in the model is also from Galveston Bay. TX, and is given in Table 5. Initially, P marinus was spec- ified to be at 50*7? prevalence in each oyster size class. This al- lowed the simulated populations to more rapidly come into equi- librium with environmental conditions than would occur using 0 or lOO'/f prevalence. The simulated distribution of P. marinus in the oyster popula- tion depends on the rate of larval recruitment and juvenile mor- tality because new recruits, being uninfected, reduce prevalence and population infection intensity. In most of the simulations, obtaining P. marinus prevalence and infection intensities that were comparable to observed values required a larval survivorship of 1 individual in 10** larvae spawned and an independent (non-P. marinus) source of juvenile mortality yielding a 1% survivorship the first year after settlement. Both survivorship rates are typical of those reported for bivalves (Brousseau et al. 1982. Powell et al. 1984. Cummins et al. 1986). Other survivorship rates were used as indicated in Table 6. MECHANISMS FOR STARTING AN EPIZOOTIC A Growing, Parasitized Oyster Population The first simulation with the oyster-parasite model was de- signed to provide a reference against which simulations consider- ing factors that produce epizootics can be compared. The reference simulation was configured to represent conditions in Galveston Bay. TX. Galveston Bay supports a substantial oyster fishery in most years and is currently in a phase of significant oyster reef expansion (Powell et al. 1995b). Food supply throughout the bay is adequate to support the present oyster population; however, a 15% decrease in food supply would restrict population growth (Powell et al. 1995a). Other environmental factors, such as tem- perature and salinity, arc usually within ranges that are conducive to oyster growth to market size (76 mm). The specific conditions used for the reference simulation are given in Tables 5 and 6. P. marinus prevalence in Galveston Bay oyster populations normally exceeds 909c (Powell et al. 1992a). and significant yearly P. mfln/iH.s-produced mortality, frequently in excess of 50% of the market-size portion of the population, can occur. How- ever, epizootics rarely occur and oyster populations normally exist in quasiequilibrium with P. marinus such that prevalence remains high and mortality remains moderate. Hence, limitations on growth of the oyster population tend to be from P. marinus- induced disease, the vagaries of larval survival, and predators such 152 Powell et al. TABLE 5. Environmental time series used as input to tlie oyster population-P. marimis model. ') Food Supply Time Senes (mg 1 ) Summer bloom (SB) — after Hofmann et al. (1 Jan Feb Mar 0.50 0.50 0.75 Summer bloom — reduced winK Jan Feb Mar 0.25 0.25 0.75 Summer bloom — reduced sumr Jan Feb Mar 0.50 0.50 0.75 Summer bloom — increased sun Jan Feb Mar 0.50 0.50 0.75 Turbidity Time Series (g 1" ') High-turbidity event (HT) Jan Feb Mar 0.00 0.00 0.00 Current Speed Time Series (cm s Low-tlow event (LF) Jan Feb Mar 1.0 1.0 1.0 Salinity Time Series (ppt) High-salinity event (Hsal) Jan Feb Mar 20 20 20 Low-salinity event (Lsal) Jan Feb Mar 20 20 20 Low-salinity event (Lrsal) Jan Feb Mar 20 20 20 Temperature Time Series (°C) Galveston Bay, Texas (GB) Given in Dekshenieks et al. (1993) High winter temperature (Ht) Winter temperatures (October to March) are 2°C Low winter temperature (Lt) Winter temperatures (October (o March) are 2°C Oyster Abundance — Confederate Reef Size (upper size limit in mm) 25, 35. 50. Abundance (number m"") 1.8 2.0 4.8 Mussel Abundance Size (upper size limit in mm) 10, 20. 30. Abundance (number m"") 50 50 50 992) Apr May June July Aug Sep Oct Nov Dec 0.75 1.25 1.25 1.25 1.25 0.75 0.75 0.50 0.50 3d (LW) Apr May Jun Jul Aug Sep Oct Nov Dec 0.75 1.25 1.25 1.25 1.25 0.75 0.75 0,25 0.25 cod (LS) Apr May Jun Jul Aug Sep Oct Nov Dec 0.75 1.00 1.00 1.00 1.00 0.75 0.75 0.25 0.25 food (HS) Apr May Jun Jul Aug Sep Oct Nov Dec 0.75 2.00 2.00 2.00 2.00 0.75 0.75 0.50 0.50 Apr May Jun Jul Aug Sep Oct Nov Dec 0.00 0.01 0.01 0.01 0.01 0.00 0.00 0.00 0.00 Apr May Jun Jul Aug Sep Oct Nov Dec 1.0 0.001 0.001 0.001 0.001 1.0 1.0 1.0 1.0 Apr May Jun Jul Aug Sep Oct Nov Dec 35 35 35 35 35 35 20 20 20 Apr May Jun Jul Aug Sep Oct Nov Dec 10 10 10 10 10 10 20 20 20 Apr May Jun Jul Aug Sep Oct Nov Dec 15 15 15 15 15 15 20 20 20 higher than those measured in Galveston Bay lower than those measured in Galveston Bay 63.5 4.6 40. 50 m. 76. 88.9 100. 110. 125. 150. !.4 1.6 2.2 0.8 0.4 0.2 0.0 )0. 60. 70. 80. 90. 100. 0 0 0 0 0 0 as crabs and oyster drills (Powell et al. 1995a). which represent top-down controls, as defined by Hunter and Price (1992). Environmental conditions that are typical of high-salinity reefs in Galveston Bay, TX. result in the simulated oyster population shown in Figure 4. A relatively stable market-size population is maintained over 5 years (Fig. 4A) and the submarket-size com- ponent (not shown) of the population rises gradually during the first 4 years and more rapidly thereafter. A decline in year 6 is produced by high population densities exceeding the available food supply. The biomass-to-length conversion given in Table I was used to calculate the number of market-size individuals. This relationship is representative of Galveston Bay oyster reefs, al- though substantial variation exists within the bay (Powell unpub- lished data). Minor mortality events, due to P. marinus, occur during the summer of the second and fourth years (Fig. 4A). The large decrease in oyster abundance seen at the end of 6 years results from the effects of crowding caused by significant popu- lation expansion in years 5 and 6 (see Powell et al. 1994 for a discussion). The oyster population maintains reproductive capa- bility throughout the simulation, with spawning occurring throughout much of the late spring and summer (Fig. 48) in each year. Triggering of Perkinsus marinus Epizootics 153 TABLE 6. The combination of environmental time series given in Table 5 and additional parameter values used for the simulations. *The different environmental time series are deFmed as: summer bloom (SB), summer bloom »ith reduced winter food (LVV), summer bloom with reduced summer food (LS), summer bloom with increased summer food (HS), high-salinity event (Hsall. low-salinitv event (Lsal), low-salinity event with slightly higher summer values (Lrsall. high-turbidity event (HTl, low -current-flow event (LFl, high winter temperatures (Ht), and low winter temperatures (Ltl. The P. marinus division time and juvenile oyster mortality used in each simulation are also shown. Except where indicated, juvenile survival was 1 in 10'*. Values indicate a 12-month continuous time series at that level. Halving Juvenile Figure Salinity Temperature Turbidity Flow Time Mortalitv Number (ppt) (°C) Food tgl"') (cm s") (hours) (d-') Comments 4 20 GB SB 0 1.0 60 0.0064 5 20 OB SB/LW /SB 0 1.0 60 0.0064 Time series split 1/1/4 6 20 GB SB/LS /SB 0 1.0 60 0.0064 Time series split 1/1/4 7 20 GB SB 0/HT /O 1.0 60 0.0064 Time series split 1/1/4 8 20 GB SB 0 1.0/LF /l.O 60 0.0064 Time series split 1/1/4 9 20 GB SB 0 1.0 60 0.0064 Mussels present days 450-650 10 20/Hsal /20 GB SB 0 1.0 60 0.0064 Time series split 1/1/4 11 20/Lsal /20 GB SB 0 1.0 60 0.0064 Time series split 1/1/4 12 20/Hsal /20 GB/Ht /GB SB 0 1,0 60 0.0064 Time series split 1/1/4 13 20 GB SB 0 1.0 120 0.0064 14 20 GB SB/LS/ HS/SB 0 1.0 60 0.0064 Time series split 1/1/1/3 15 20/Hsal/ . Lrsal/20 GB/Ht/ LT/HB SB 0 1.0 60 0.0064 Time series split 1/1/1/3 16 20 GB SB 0 1.0 60 0.0064 Summer recruitment rate; year 2. 1 in 10""; year 3, 6 in IC One of the checks on the simulation is to ensure that the sim- ulated seasonal progression of P. marinus infection intensity and prevalence corresponds to measured patterns. The observed pat- tern of P. marinus prevalence in oyster populations usually shows lows in late winter to early spring, an increase in late spring, and a peak in mid to late summer. The pattern of prevalence for the total simulated oyster population (Fig. 4C, solid line) shows lows in the summer and highs in the winter, which is exactly the op- posite of field measurements. The lows in prevalence in the sim- ulated populations occur during recruitment following major spawning events. The standard approach for measuring P. mariinis prevalence involves the collection of the largest individuals in the population, normally those of market size. If only this portion of the simulated oyster population is considered, prevalence exceeds 80% in most months of the year (Fig. 4C, dotted line). Lows occur in fall and winter as submarket-size adults grow to market size, but preva- lence does not decline to the normally measured winter levels. Recent experimental studies (Choi et al. 1989) have shown that the thioglycollate technique typically used to assess P. marinus prev- alence (Ray 1966) frequently misdiagnoses light infections as neg- ative. Thus, if it is assumed that infections of s2'- cells ind~ ' are normally misdiagnosed as negative, then the pattern of prevalence in the market-sized portion of the simulated oyster populations shows the observed seasonal cycle (Fig. 4C. dashed line). Low prevalences occur in February and March, when many false negatives are reported, and peak prevalences occur in the late summer. A similar problem exists for the calculation of mean infection intensity for the population depicted in Figure 4B when using Mackin's (1962) Scale as modified by Craig et al. (1989). The simulated seasonal progression of infection intensity matches the observed pattern only when the market-sized portion of the pop- ulation is considered (Fig. 4B. dotted line). The seasonal cycle of infection intensity in the entire population (Fig. 4B. solid line) can be discerned but summer highs are depressed as disease intensifi- cation in the adults is offset by recruitment of uninfected individ- uals. Thus, the seasonal cycle that is observed in P. marinus prevalence and infection intensity is dependent upon the size class structure of the sampled population. The routine sampling of only the largest individuals in the population normally does not accu- rately portray the disease status of the entire population. The im- plications of this are discussed more fully in Hofmann et al. (1995). The level off. marinus prevalence that occurs in the simulated oyster populations shown in Figure 4 in response to Galveston Bay environmental conditions remains above 609c throughout most of the year. Yearly highs exceed 90% in most years. Mean infection intensity of the market-size adults reaches 4 on Mackin's Scale in years when mortality occurs. Infection intensity in the entire oyster 154 Powell et al. 20 j /zzyi 00 80 i^H 60 1 40 - 1 20 Sl iJgL -^^ U — ^^^ — ^*- — — ! 1 1 23456789 10 11 Size Class Figure 3. Size-frequency distribution of oysters that was used to ini- tialize the oyster population model. Data are from observations made at South Deer Island in the West Bay section of Galveston Bay, TX, in spring 1992. population averages 2 to 3. a light to moderate infection, during the summer and fall. This is lower than that for the market-size population due to the dilution effect of new recruits. The infection intensity for the entire oyster population drops to about 1 during the winter. This is higher than the infection level in the market-size fraction of the populations because the newly recruited smaller individuals, with the same number of P. marinus cells, have higher infection intensities on a cell per gram basis. Infection intensity in the market-size individuals is about 3.5 on Mackin's Scale (moderate infections) in most years. However, in years in which there is significant P. marinus mortality, infec- tion intensity nears 4 on Mackin's Scale. This small variation in infection intensity, which corresponds to about 1 to 2 population doublings (Hofmann et al. 1995). is all that is required to separate years of moderate-to-low mortality from years having significant mortality events. TRIGGERING MECHANISMS FOR EPIZOOTICS Rapid growth and high fecundity are the principal defenses against predation and disease for many host-prey/parasite-predator systems (e.g.. Onstad and Maddox 1989. Warburton 1958). Oys- ter populations are no exception, with recruitment, growth, and fecundity usually exceeding, by some small amount, the combined rate of P. marinus transmission, intensification, and mortality. Therefore, mechanisms that can potentially trigger epizootics should be sought primarily among the variables that modify the oyster population potential for recruitment, growth, and fecundity. Obvious choices for potential triggering mechanisms are variations in environmental conditions, such as food supply, turbidity level, current flow, salinity, and temperature. Other factors that influ- ence the ability of the oyster population to grow such as compe- tition for food (i.e., mussels), variations in recruitment and juve- nile mortality, and the ability of the oyster to resist disease also potentially affect the occurrence of epizootics. Frequently, the factorfs) that triggers an epizootic occurs well before the detection of the event (Gill 1928). and once initiated, epizootics can persist during what would be considered normal or optimal conditions. Furthermore, only a small change in condi- tions may be needed to trigger an epizootic because populations often exist in quasiequilibriuni with the disease (Anderson 1991, Lenski and May 1994). Thus, the simulations that were designed to investigate epizootic triggering mechanisms used food, temper- -1.0 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I A JMMJSNJMMJSNJMMJSNJMMJSN Year 1 Year 2 Year 3 Year 4 Time JMM J SN Years JMM JSNJ Year 6 Triggering of Perkinsus marinus Epizootics 155 ature, salinity, and turbidity conditions for Galveston Bay in u Inch 1 year of the 6-ycar time series was modified to introduce a small change in conditions. For each simulation, year 1 repre- sented noniial environmental conditions (as used for the simula- tion shown in Fig. 4), year 2 included the modified condition, and vears 3 to 6 returned to normal conditions. Thus, the oyster pop- ulations were exposed to 5 years of environmental conditions that are conducive to growth and expansion (Fig. 4) and I year that potentially was not. Food Supply Decreased food supply reduces oyster growth and fecundity (Soniat and Ray 1985, Robinson 1992) but does not affect the cell division rate of P. inariniis. Low food supply, then, is potentially an epizootic-triggering mechanism. However, the time during which oysters experience low food supply can be important because the rate of P. marinus division is temperature dependent. Thus, the effect of low food supply might be expected to be less in the winter than in the summer. In the following simulations, food supply in the second year was reduced by 0.25 mg 1 " ' for 4 months in either the winter or the summer, which gives a 109i: decrease in food over the year. A reduction in food supply during the winter has little impact on the oyster population (Fig. 5). Oyster ingestion rates are pri- marily a function of filtration rate which is temperature controlled. Decreased temperatures in the winter result in reduced filtration rates so that the impact of low food supply during this time is minimal. Moreover, the division rate of P. marinus is at its yearly low. Thus, low winter food supplies do not substantially alter the pattern of P. marinus prevalence and infection intensity from that seen in the reference simulation. P. marinus mortality is increased somewhat, but the oyster population continues to grow and ex- pand. By contrast, a reduction in food supply during the summer triggers an epizootic (Fig. 6). This epizootic contains all of the basic characteristics of most epizootics (e.g.. Gill 1928, Plowright 1982. Shields and Kuris 19881. Although the reduced food supply occurs in the summer of the second year, the response of the oyster population is not immediately obvious and no dramatic mortality event occurs in the following year (year 3, Fig. 6A). In the next 2 years (4 and 5). however, the population declines to extinction as the primary phase of mortality begins 18 months after the trigger- ing event (Fig. 6A). Spawning continues during the entire epizootic phase and rates do not decline substantially until significant mortality begins in the adult population in year 5 (Fig. 6B1. Fecundity through year 5 would be adequate for population recovery were P. nuirinus in- fection intensity to decline. Infection intensity rises persistently from about 3 in the summer and 1.5 in the winter of year 2 to above 4 in the summer and 2 in the winter in year 6 (Fig. 6B). The rise in infection intensity is most noticeable in the entire popula- 1 I I I I I I I I II I I I I I I I II I I I I I I II I I I I I II I I I I I I II I I I I I I I I I I I I I I I I I JMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJ Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Time Figure 5. The number of market-size individuals (solid linel in the population expressed as log,„(number m~~), from a simulation that used environmental conditions that are characteristic of a high-salinity reef in (iaiveston Bay, TX (Table 5), that experienced a decline in food supply during the v\inter of year 2 (Table 6). Mortality events, calcu- lated as the fraction of the population in a given size class that dies during a 1-month period, are indicated by the shaded contours, with an interval of 0.1. tion rather than in the normally sampled market-size component because mortality in the latter size classes continually removes the most heavily infected individuals from the population. During the epizootic, disease prevalence in the population gradually increases from about 60 to 80% to near 100% (Fig. 6C). However, prevalence, as usually measured (dotted line in Fig. 6C). changes little during the epizootic. Thus, false-negatives and inadequate sampling of the entire oyster size-frequency distribu- tion can inhibit observation of this phase of disease intensification. In this simulation, the population crash is not produced by a dramatic decrease in fecundity or recruitment. These are products of the crash. The population crash occurs because the rate of P. marinus growth and transmission exceeded the rate of expansion of the oyster population by just a small amount in year 2. A reduction in food supply during the summer months produces a subtle change in the balance between oyster population expansion and disease intensification which permits the disease to nudge ahead and gradually exert control over the host population. The initial food conditions for this simulation, which are typical of Galveston Bay. TX. allow population expansion but are near the threshold that can trigger an epizootic. For higher food supplies, a 10% decrease would have had a lesser effect. Thus, this simulation shows that a small change in environmental conditions may be all that is needed to generate an epizootic once the population nears the carrying capacity of the environment, and once the epizootic is triggered, simply returning to pretrigger environmental conditions Figure 4, (Al The number of market-size individuals (solid linel in the population expressed as log,„(number m"'), from a simulation that used environmental conditions that are characteristic of a high-salinity reef in (iaiveston Bay, TX (Table 5). Mortality events, calculated as the fraction of the population in a given size class that dies during a I -month period, are indicated by the shaded contours, ^^ith an interval of 0.1. (Bl Simulated oyster population reproductive eflort (shading) expressed as log|,|(lotal joules spawned per month) in each size class. Contour interval is 2 log units. P. marinus infection intensity expressed in terms of Mackin's (19621 0- to 5-point scale is shown for the entire population (solid line) and the market-size (*3-inch) portion of the population (dotted line). (C) P. marinus prevalence expressed as the fraction of the total population that is infected. Prevalences in the entire oyster population, the market-size (s:3-inch) portion of the oyster population, and the market-size population, assuming that all infections =s2'- cells ind ' are judged negative using the method described by Ray (1966), are represented by the solid, dotted, and dashed lines, respectively. 156 Powell et al. is insufficient to prevent disaster. Moreover, the simulation shows that an epizootic generated by a small but significant decline in food supply can be characterized by a delay of about 18 months between the trigger and an observed increase in mortality and that a decline in fecundity may not become obvious until significant mortality begins in the adult population. An increase in prevalence and infection intensity in the population may serve as an early warning sign of an impending epizootic. However, this rise may only be noticeable in that fraction of the population smaller than market size, a fraction normally not sampled by field surveys. Turbidity Increased turbidity decreases feeding efficiency and therefore should also restrict food supply. Increasing turbidity during the summer of year 2 to 10 mg P ' (Table 5) initiates an epizootic that produces significant mortality about 12 months after the high- turbidity event and results in a crash of the oyster population in year 4 (Fig. 7). The simulated disease prevalence and intensity associated with this event are essentially identical to those ob- tained for the reduced food scenario (Fig. 6). Current Flow Certain combinations of food content, population density, and current flow may significantly affect the flux of food over the oyster reef (Muschenheim 1987. Wilson-Ormond et al. in press). A reduction in current flow during the four summer months of year 2 gives results that are similar to those shown for reduced food conditions. Significant oyster mortality begins about 18 months after the low-flow event and the population eventually crashes in years 5 and 6 (Fig. 8). The pattern of intensification of P. marinus infection in this and the reduced food simulation is similar, but the epizootic that results from low current flow develops more slowly. Mussels Competition from other filter feeders may reduce food supply and thus adversely impact the oyster population. In Texas bays, mussels of the genus Brachidontes are abundant. To provide a simulation comparable to the low-food, increased turbidity, and low-flow simulations, the mussels were allowed to impact food supply for only the summer months in year 2. The competing effect of the mussels acts to decrease the food supply to the oys- ters. The time-dependent evolution of the oyster population (Fig. 9) is similar to that shown in Figure 6 and the pattern of disease intensification and intensity is essentially identical to that seen in Figure 6B and C. An epizootic triggered in year 2 results in sig- nificant mortality about 18 months later and the population begins to decline in years 4 and 5 (Fig. 9). Salinity Small changes in climate have been shown to significantly modify recruitment to marine populations (Turrell et al. 1992) and disease (Jarosz and Burden 1992). P. marinus responds to tem- perature and salinity variations and even small perturbations in these environmental conditions arising from changes in climate can significantly modify P. marinus disease intensity over large geographic areas (Powell et al. i992a). High-salinity conditions have a greater impact during the warmer half of the year. Tem- perature has a major effect in winter through its influence on parasite division rate and mortality. The effect of salinity was investigated by raising salinity by 15 I I I I I I I I I I I I I I I I I I I I I I I I I [ I I I I MM A TTTTTTTTTTTTT JMUJSNJMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJ Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Time Triggering of Perk/nsus marinus Epizootics 157 I I M I I M I I I I I I I I I I I I I I I I I JMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJ Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Figure 7. The number of market-size individuals (solid line) in the population expressed as log,„(number m -), from a simulation that used environmental conditions that are characteristic of a high-salinity reef in Galveston Bay. TX (Table 5), that experienced an increase in turbidity during the summer of year 2 (Table 6). Mortality events, calculated as the fraction of the population in a given size class that dies during a 1-month period, are indicated by the shaded contours, with an interval of 0.1. I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I .2.0 I I I I I I [ 1 I I I I r I 1 I r I I r I ] I I I I I I 1 I I JMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJ Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Time Figure 8. The number of market-size individuals (solid linel in the population expressed as log||,(number m 'l, from a simulation that used environmental conditions that are characteristic of a high-salinity reef in Galveston Bay, TX (Table 5), that experienced a decrease in current flow during the summer of year 2 (Table 6). Mortality events, calculated as the fraction of the population in a given size class that dies during a 1-month period, are indicated by the shaded contours, with an interval of 0.1. ppt for 6 months. April to September, in year 2. Increased salinity during the warmer months reduces the rate of population growth in comparison to the simulation that used normal conditions (cf. Fig. 4) by increasing the mortality of market-size oysters from P mari- nus. However, an epizootic does not occur (Fig. 10). High-salinity is usually associated with epizootics (e.g., Crosby and Roberts 1990. Mann et al. 1991). However, many of the oyster popula- tions in the Gulf of Mexico e.xist at salinities above 20 ppt for much or all of the year and maintain productive and expanding populations. High salinity may facilitate the development of an epizootic, but high salinity alone is unlikely to trigger one. Exposing an oyster population to a 10-ppt decrease in salinity for 6 months, however, produces an immediate mortality event, which continues for an indefinite time (Fig. IIA). During the low-salinity event, infection intensity (Fig. IIB) and prevalence (Fig. IIC). as usually measured, decline as expected. However, the population prevalence and infection intensity rise as individu- als of market-size decline, which is counterintuitive. Low salinity restricts scope for growth, and in the absence of a balancing effect such as increased food supply, this decrease restricts oyster growth, particularly in individuals already growth restricted by high P marinus infection intensity (e.g., Menzel and Hopkins 1955). Such a growth restriction would be just enough, in heavily infected oysters, to produce a lethal infection. It is well known that oysters are more sensitive to low-salinity mortality during the sum- mer months (e.g., Gunter 1953. Ray 1987. E. Powell unpublished data). This simulation suggests that P. mannus infection may be one important reason for this sensitivity, although no observations are available to support this speculation. Furthermore, the dra- matic decrease in prevalence and infection intensity noted during and after low-salinity events in the summer (e.g.. Soniat 1985) may well be due as much to the removal of heavily infected in- dividuals from the population as to inhibition of P . marinus in- tensification. No evidence from field observations is available to support or refute this suggestion. Temperature Water temperature variation in Gulf of Mexico bays and estu- aries between warm and cold years is rarely more than 2°C from the long-term mean (Sittel 1994). A change in temperature of this magnitude failed to initiate an epizootic. Frequently, however, extremely warm years co-occur with extremely dry years (about 107f of all years) in Galveston Bay and these years may be char- acterized by relatively warm winters or relatively warm summers. To test these effects, the summer temperature for the Galveston Bay time series was increased by 2°C (Table 5) and used with the summer salinity conditions that produced the simulated oyster population shown in Figure 10. The resultant extremely warm and dry summer produced a simulated oyster population distribution that was not significantly different from that shown in Figure 10. One reason is that summer conditions in the Gulf of Mexico are Figure 6. (.\l The number of market-size individuals (solid line) in the population expressed as log,|,(number m -|, from a simulation that used environmental conditions that are characteristic of a high-salinity reef in (Jalveston Bay, TX (Table 5), that experienced a decline in food supply during the summer of year 2 (Table 6). Mortality events, calculated as the fraction of the population in a given size class that dies during a 1-month period, are indicated by the shaded contours, with an interval of 0.1. (Bl .Simulated oyster population reproductive effort (shading! expressed as logmttotal calories spawned per monthl. P. marinus infection intensity expressed in terms of Mackin's ( 1962) 0- to 5-point scale is shown for the entire population (solid line) and the market-size (s3-inch) portion of the population (dotted linel. (C) P. marinus prevalence expressed as the fraction of the total population that is infected. Prevalences in the entire oyster population, the market-size (&3-inch) portion of the oyster population, and the market-size population, assuming that all infections !£2'" cells ind"' are judged negative using the method described by Ray (1966), are represented by the solid, dotted, and dashed lines, respectively. 158 Powell et al. 2.0 1.0 -1.0 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I r I r I I I I I JUMJSNJMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJ Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Time Figure 9. The number of market-size individuals (solid line) in the population expressed as log,„(number m~'), from a simulation that used environmental conditions that are characteristic of a high-salinity reef in Galveston Bay, TX (Table 5), that experienced a competitive interaction with mussels during .lulian Days 450 to 630 (Table 6). Mortality events, calculated as the fraction of the population in a given size class that dies during a I -month period, are indicated by the shaded contours, with an interval of 0.1. 4.0 - 2.0 - 0.0 -1.0 I I I I I I I I I I I I I I I I I I M M I I I I I I I I I I I I I r I t I I I I I I I I I I I I I I JMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJ Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Figure 10. The number of market-size individuals (solid line) in the population expressed as log,„(number m~~), from a simulation that used environmental conditions that are characteristic of a high-salinity reef in Galveston Bay, TX (Table 5), that experienced a high-salinity event during the summer of year 2 (Table 6). Mortality events, calcu- lated as the fraction of the population in a given size class that dies during a 1-month period, are indicated by the shaded contours, with an interval of 0.1. already so conducive to P. marinus intensification that slightly warmer conditions have very little additional (nipact. Oyster pop- ulations must routinely withstand warm, dry summers to maintain the population abundances normally observed. The same is not the case for a warmer winter. For this simu- lation, the winter water temperatures from Galveston Bay were increased by 2°C (Table 5) and used with the higher summer salinity conditions (Fig. 10) to produce a dry summer following a warm winter. These conditions produce a classic P . marinus epi- zootic (Fig. 12) which is similar in all respects to those described in previous simulations (e.g.. Figs. 6-8). A warm winter increases the ratio of parasite division rate to parasite mortality rate so that winter infection intensities remain relatively high. This simulation suggests that the coincidences of appropriate summer salinities and winter temperatures arc the environmental factors that contribute the most to the generation of an epizootic in Gulf of Mexico bays and estuaries. Recruitment and Juvenile Mortality Factors that affect population fecundity, recruitment, or juve- nile mortality may destabilize host/parasite populations in quasiequilibrium (Dobson 1988). Decreasing recruitment success by 50Vc in the summer of year 2 (Julian days 450 to 630) or increasing juvenile mortality by 50% in the same time frame pro- duced an epizootic qualitatively identical to the one depicted in Figure 6. The intensification of P. marinus infection closely fol- lowed the pattern shown in Figure 6B and C. An epizootic began to produce significant mortality about 1 2 months after the event, in each case, and the population crashed in years 4 and 5. Changing Disease Resistance or Virulence Resistance to disease is often important in initiating or stopping an epizootic (Ross 1982. Kent et al. 1989. McCallum 1990, Moller 1990). Although the development of resistance to P. marinus has been questioned (e.g., Lewis et al. 1992), Hofmann et al. (1995) suggested that sonic regional variation in oyster resistance or P. mari- nus virulence is probably required to explain regional variations in P. minimis prevalence and infection intensity. This effect was simulated by reducing the rate of parasite mortality ('■,„(7',S)| by changing the population halving time at 20°C-20 ppt from 60 to 120 hours. Note that a reduction in population doubling time (increased virulence) would yield similar results. The resultant simulated oyster population undergoes an epizootic with mass mortality starting in year 3 (Fig. 13). The reduction in parasite mortality (or increase in parasite divi- sion time) primarily affects the winter drop in infection intensity and produces conditions similar to those produced by a warm winter. Vanations in the rate of parasite division or mortality have little effect in the summer when parasite density effects exert a major control on the growth rate of the P . marinus population. MECHANISMS FOR STOPPING AN EPIZOOTIC General Considerations Once started, an epizootic is difficult to stop. In most cases, epizootics cease when the host population's density drops to a Figure 11. (A) The number of market-size individuals (solid line) in the population expressed as log,„(number m ~), from a simulation that used environmental conditions that are characteristic of a high-salinity reef in Galveston Bay, TX (Table 5), that experienced a low-salinity event during the summer of year 2 (Table 6). Mortality events, calculated as the fraction of the population in a given size class that dies during a I-month period, are indicated by the shaded contours, with an interval of 0.1. (B) P. marinus infection intensity expressed in terms of Mackin's (1962) 0- to 5-point scale is shown for the entire population (solid line) and the market-size (5=3-inch) portion of the population (dotted line). (C) P. marinus prevalence expre.ssed as the fraction of the total population that is infected. Prevalences in the entire oyster population, the market-size (3=3-inch) portion of the oyster population, and the market-size population, assuming that all infections =£2'" cells ind ' are judged negative using the method described by Ray (1966), are represented by the solid, dotted, and dashed lines, respectively. Triggering of Perkinsus marinus Epizootics 159 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I A 9.0 8.0 - 3.0 I I M I M I I I 3.0 IJI I I I I I I I I 1 I 11 ' I I I I I I ■ I I I ' I I ' - 3.0 JUMJSNJMMJSNJMMJSNJMMJSNJMMJSNJMMJSNJ Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Figure 12. The number of market-size individual.s (solid line) in the population expressed as log|„(number m "), from a simulation that used environmental conditions that are characteristic of a high-salinity reef in (Jalveslon Bay, TX (Table 5», that experienced a high-salinity event during the summer and a warm temperature event during the \2-299. Mackin, J. G. & J. L. Boswell. 1954. The relation of forced closure of oysters and acceleration of reproduction of Dermocystidium marinum. Te.xas A&M Univ. Res. Fouiul. Tech. Rep. Proj. 23 17:1-6. Mackin. J. G. & S. H. Hopkins. 1961. Studies on oyster mortality in relation to natural environments and to oil fields in Louisiana. Publ. Inst. Mar. Sci. Univ. Te.xas 7:1-131. Mackin. J. G. & S. M. Ray. 1955. Studies on the effect of infection by Dermocystidium marinum on ciliary action in oysters (Crassosirea vir- ginica). Proc. Natl. Shellfish. Assoc. 45:168-181. Mann. R.. E. M. Burreson & P. K. Baker. 1991. The decline of the Virginia oyster fishery in Chesapeake Bay: Considerations for intro- duction of a non-endemic species, Crassostrea gigas (Thunberg, 1793). J. Shellfish Res. 10:379-388. May, E. B. 1971. A survey of the oyster and oyster shell resources of Alabama. Ala Mar. Res. Bull. 4:1-53. McCallum, H. I. 1490. Covariance in parasite burdens: the effect of pre- disposition to inloction. Parasitology 100:153-159. McCallum, H, I, & G R. Singleton. 1989. Models to assess the potential of Capillarui hepatica to control population outbreaks of house mice. Parasitology 9^:-\2i-i,il . Medcof. J. C. 1961. Oyster farming in the maritimes. Bull. Fish. Res. Board Can. 131:1-155. Menzel. R. W. & S. H. Hopkins. 1955. The growth of oysters parasitized by the fungus Dermocystidium marinum and by the trematode Bu- cephalus cuculus. J. Parasitol. 41:333-342. Moller, H. 1990. Association between diseases of flounder {Ptatichthys flesus) and environmental conditions in the Elbe estuary, ERG. J. Cons. Int. E.xplor. Mer 46:187-199. Mollison, D. 1987. Population dynamics of mammalian diseases. Symp. Zoo/. Soc. Land. 58:329-342. Muschenheim, D. K. 1987. The dynamics of near-bed seston flux and suspension-feeding benthos. J. Mar. Res. 45:473-496. Newell, R. I. E. 1985. Physiological effects of the MSX parasite Haplo- sporidiuni nelsoni (Haskin, Stauber & Mackin) on the American oyster Crassostrea virginica (Gmelin). J. Shellfish Res. 5:91-95. NOAA. 1991. Recreational shcllfishing in the United States. United States Department of Commerce. National Oceanic and Atmospheric Admin- istration. 22 pp. Ogle. J. & K. Flurry. 1980. Occurrence and seasonality of Perkinsus marinus (Protozoa: Apicomplexa) in Mississippi oysters. Gulf Res. Rep. 6:423-425. Onstad. D. W. & J. V. Maddox. 1989. Modeling the effects of the mi- crosporidium. Nosema pyrausta. on the population dynamics of the insect. Ostrinia nubilalis. J. Invertebr. Pathol. 53:410-421. Paynter. K. T. & E. M. Burreson. 1991. Effects of Perkinsus marinus infection in the Eastern oyster, Cras.sostrea virginica: II. Disease de- velopment and impact on growth rate at different salinities. J. Shellfish Res. 10:425-431 Plowright, W. 1982. The effects of rinderpest and rinderpest control on wildlife in Africa. Symp. Zool. Soc. Loud. 50:1-28. Powell, E. N., H. Cummins, R. J. Stanton, Jr. & G. Staff. 1984. Esti- mation of the size of moUuscan larval settlement using the death as- semblage. Estuarine Coastal Shelf Sci. 18:367-384. Powell, E. N., J. D. Gauthier, E. A. Wilson, A. Nelson, R. R. Fay & J. M. Brooks. 1992a. Oyster disease and climate change. Are yearly changes in Perkinsus marinus parasitism in oysters (Crassostrea vir- ginica) controlled by climatic cycles in the Gulf of Mexico? P.S.Z.N.I.:Mar. Ecol. 13:243-270. Powell, E. N., E. E. Hofmann, J. M. Klinck & S. M. Ray. 1992b. Mod- eling oyster populations I. A commentary on filtration rate. Is faster always better? J. Shellfish Res. 1 1:387-398, Powell, E. N., E. E. Hofmann, J. M. Klinck, E. Wilson-Omiond & M. S. Ellis. 1995a. Modeling oyster populations V. Declining phyto- plankton stocks and the population dynamics of American oyster (Crassostrea virginica) populations. Fish. Res. 24:199-222. Powell, E. N., J, M, Klinck, E, E. Hofmann & S. M. Ray. 1994, Mod- eling oyster populations. IV. Population crashes and management. U. S. Fish Wildl. Sen-. Fish. Bull. 92:347-373. Powell, E. N., J. Song, M. S. Ellis & E. A.Wilson-Omiond. 1995b. The status and long-term trends of oyster reefs in Galveston Bay, Texas. J. Shellfish Res. 14:439-457. Powell, E. N. & R. J. Stanton, Jr. 1985. Estimating bioniass and energy flow of molluscs in paleo-communities. Palaeontology (LonJ) 28:1- 34. Quick, J. A., Jr. & J. G. Mackin. 1971. Oyster parasitism by Labyrinth- omy.xa marina in Florida. Fla. Dept. Nat. Res. Mar. Res. Lab. Prof. Paper Ser. 13:1-55. Ragone, L. M. & E. M. Burreson. 1993. Effect of salinity on infection progression and pathogenicity of Perkinsus marinus in the Eastern oyster, Crassostrea virginicia (Gmelin). J Shellfish Res. 12:1-7, Ray. S, M, 1954, Biological studies of Dcrmoc y.sri(/mm munnHm a fungus parasite of oysters. Rice Inst. Pamph. Monogr. Biol. Spec. Issue 114 pp. Ray. S. M. 1966. A review of the culture method for detecting Dermo- Triggering of Pf;RKiNsus marinus Epizootics 165 cysiidiiim nuiniuim. with bUggested modifications and precaulions. Proc. Natl. Shellfish. Assoc. 54:55-69. Ray, S. M. 1987. Salinity requirements of the American oyster. Crassos- trea virginicia. In: A. J. Muller & G. A. Matthews (eds.). Freshwater Inflow Needs of the Matagorda Bax System with Focus on the Needs of Penaeid Shrimp. National Oceanic and Atmospheric Administration, Technical Memorandum NMFS-SEFC-189. pp. E.I-E.28. Ray, S. M. & A. C. Chandler. 1955. Dermocxstidiiim marinum. a parasite of oysters. E.xp. Parasilol. 4:172-200, Robinson, A. 1992. Dietary supplements for reproductive conditioning of Crassostrea gigas kiimamoto (Thunberg). I. effects on gonadal devel- opment, quality of ova and larvae through metamorphosis. J. Shellfish Res. 11:437-441 Ross, J. 1982. Myxomatosis: the natural evolution of the disease. S\mp. Zool. Soc. London 50:77-95 Saunders, G. L., E. N. Powell & D. H. Lewis. 1993. A determination of in vivo growth rates for Perkinsiis marinus. a parasite of Crassostrea virginica. J. Shellfish Res. 12:229-240. Schmid-Hempel, P. & R. Schmid-Hempel. 1988. Parasitic flies (Conop- idae, Diptera) may be important stress factors for the ergonomics of their bumblebee hosts. Ecol. Entomol. 13:469-472. Shields, J. D. & A. M. Kuris. 1988. Temporal variation in abundance of the egg predator Carcinoneinertes epialii (Nemertea) and its effect on egg mortality of its host, the shore crab, Hemigrapsiis oregonensis. Hydrobiologia 156:31-38. Sindermann, C. J. 1993. Disease nsks associated with imponaiion of non- indigenous manne animals Mar. Fish. Rev. 54:1-10. Sittel, M. C. 1994. Marginal probabilities of the extremes of ENSO events for temperature and precipitation in the southeastern United States. Florida State University Technical Report, 94-1, 155 pp. Soniat, T. M. 1985. Changes in levels of infection of oysters by Perkinsus marinus. with special reference to the interaction of temperature and salinity upon parasitism. Northeast Gulf Sci. 7:171-174. Soniat, T. M. & M. S. Brody. 1988. Field validation of a habitat suit- ability index model for the American oyster. Estuaries 11:87-95. Soniat, T. M. & S, M. Ray. 1985. Relationships between possible avail- able food and the composition, condition and reproductive state of oysters from Galveston Bay, Texas. Contrib. Mar. Sci. 28:109-121. Turrell, W R 1992. New hypotheses concerning the circulation of the northern North Sea and its relation to North Sea fish stock recruitment. ICES J Mar. Sci. 49:107-123. Warburton, F. E. 1958. Control of the bonng sponge on oyster beds. Fish. Res. Bd. Can. Prog. Rep. Atlantic Coast Stat. 69:7-1 1. Wells, H. W. 1961 . The fauna of oyster beds, with special reference to the salinity factor. Ecol. Monogr. 31:239-266. White, M. E.. E. N. Powell & S. M. Ray. 1988. Effects of parasitism by the pyramidellid gastropod Boonea impressa on the net productivity of oysters iCrassoslrea virginica). Estuarine Coastal Shelf Sci. 26:359- 377. Wilson. E. A., E. N. Powell, M. A. Craig, T. L. Wade & J. M. Brooks. 1990. The distnbution o( Perkinsus marinus in Gulf coast oysters: its relationship with temperature, reproduction, and pollutant body bur- den Int. Rev. Gesamlen Hydrobiol. 75:533-550. Wilson-Ormond, E. A., E. N. Powell & S. M. Ray. in press. Short-term and small-scale variation in food availability to natural oyster popula- tions: food, flow and flux. P. S.Z.N. I.: Mar. Ecol. Wright, D. A. & E. W. Hetzel. 1985. Use of RNA:DNA ratios as an indicator of nutritional stress in the American oyster Crassostrea vir- ginica. Mar. Ecol. Progr. Ser. 25:199-206. Journul of ShflljUh Research. Vol. 15. No. 1, 167-176. 1996. MANAGEMENT ALTERNATIVES FOR PROTECTING CRASSOSTREA VIRGINICA FISHERIES IN PERKINSUS MARINUS ENZOOTIC AND EPIZOOTIC AREAS G. E. KRANTZ AND S. J. JORDAN Manlaiul Deparimeni of Natural Resources Cooperative Oxford Laboratory 904 S. Morris St. Oxford. Maryland 21654 ABSTRACT Wc review management of oyster stocks infected by Perkin.sKs mcinmis. comparing previously published recommen- dations with current practices, with emphasis on the public fishery in the Maryland portion of Chesapeake Bay. The epizootiology of perkinsiasis is descnbed. particularly the spread of the disease into low salinity areas. We also describe recent attempts to develop a policy and management framework for restoration of oyster populations that have been depleted by P manims and Haplosporidiiim nelsoni. It is apparent from experiences in Gulf of Mexico estuaries. Long Island Sound, and Maryland that strong recruitment can. to some extent, offset the impacts of P. marimis on oyster fisheries. Although improved management practices so far have had very limited success in maintaining harvestable stocks in the Chesapeake, it is clear that the recruitment potential of oyster populations has not been diminished to a critical point. Strategies designed to enhance and supplement natural recruitment, along with maintaining growing areas as free from P. mahnus infections as possible, currently offer the most promise for maintaining harvestable stocks. In combination, new developments in research, management, monitonng. and policy are cause for guarded optimism, both for larger, sustainable harvests and for restoration of some of the ecological functions of healthy oyster populations. KEY WORDS: Cras.sosrrea virginica. Perkinsii.'i mariniii. oyster management, fisheries management INTRODUCTION The protozoan Perkinsus mahnus. phylum Apicomplexa, par- asitizes eastern oysters. Crassostrea virginica, causing extensive mortality throughout its ecological range. The present known range for the parasite extends from the southern Gulf of Mexico through southern New England. Basic aspects of the taxonomy, life history, and ecology of the parasite have been described (An- drews 1954. 1965. 1988; Mackin 1962: Ray 1966; Perkins 1988). Many field and laboratory infection studies have provided suffi- cient information to predict the impacts of enzootic and epizootic episodes of perkinsiasis ("dermo disease") in oyster populations (e.g.. Paynter and Burreson 1991. Fisher et al. 1992. Smith and Jordan 1992). Two of the more prolific investigators of this dis- ease. Andrews and Ray (1988) delineated several management strategies to control the impacts off. marinus in oysters. Their suggested management practices focused on manipulation of pri- vately owned or leased beds where oyster populations were tradi- tionally completely harvested and replaced by a new crop of seed oysters. However, several important attributes of the host-parasite interaction need to be reconsidered and some of the basic man- agement concepts modified to reflect recent research findings (Saunders et al. 1993. Bushek et al. 1994. Roberson et al. 1993). In this paper, we first review the interactions of the host-parasite relationship with 1) environmental factors. 2) oyster recruitment, and 3) traditional fishery and management practices. Second, we reiterate long-established concepts of how oyster populations can be managed to minimize the impacts of perkinsiasis. Third, we present the current status of management and some future direc- tions, with emphasis on the experience in Maryland and the Ches- apeake Bay. INFLUENCE OF ENVIRONMENTAL FACTORS ON P. MARINUS EPIZOOTIOLOGY Throughout its range. P. marinus is most active in producing pathology in oysters during the warmer months of the year. Per- kinsiasis exhibits a gradient of infection with high prevalence and intensity in high salinity areas and lower infection levels in pop- ulations at low salinity in the same estuaries. North of Cape Hat- teras. the parasite becomes dormant during winter and early spring and frequently is undetectable by histology or by thioglycollate culture of rectal and gill tissue. Subpatent infections are known to exist in many individuals within an overwintering population. These animals initiate new generations of intense infections in the population from May through October. The process of re- establishment of detectable infections in individual oysters is com- promised by water temperatures below 20°C and by salinity below 8-10 ppt. As a result, a gradient of severe to light infections is re-established annually in populations along the salinity gradient in all estuaries. Upon re-establishment of detectable infections in spring, a range in diagnosable infections may be found even in populations that had 1007f high intensity infections in the fall of the previous year. The gradient of P. marinus disease prevalence and intensity in estuaries was first reported by Mackin (1962) in his classic study where he found heavy infections of P. marinus in oyster populations living in environments where salinity fre- quently dropped to 2 ppt. Surveys along the coast of the Gulf of Mexico conducted by later investigators (Craig et al. 1989. Soniat 1985. Turner 1985) described variations in disease pressure among oyster populations and continued the existence of active P. mari- nus infections in oyster populations living in salinities of 1-2 ppt. Since 1988. perkinsiasis has become more active in Chesa- peake Bay and has spread into the low salinity environments of Maryland and Virginia estuaries. The disease is now found throughout oyster-producing areas of the Chesapeake. Populations that have high prevalence and intensity of infection during the summer months have high annual mortality regardless of their geographic location (Andrews and Ray 1988. Otto and Krantz 1980. Krantz. 1991, 1993, 1995). Infections by P. marinus in Chesapeake Bay oysters occur primarily in the warm summer months; previous investigators hy- pothesized that direct transmission of P. marinus occurred from 167 168 Krantz and Jordan overwintered infected oysters to uninfected oysters and that prox- imity of uninfected to infected oysters was a significant variable subject to management manipulation. Careful monitoring of the water column using optical sorting of filtered particles labeled by fluorescent antibodies has suggested that P. marimis may be widely spread throughout the Chesapeake Bay by watcrborne stages of the parasite (Roberson et al. 1993). Oyster exposure studies in trays deployed in Chesapeake Bay demonstrated that site-specific infection rates had more influence on seasonal levels of P. marinus than initial mfection levels in the experimental oysters (Meritt 1993). In an unpublished study. G. F. Smith (Co- operative Oxford Laboratory) found evidence that large doses of apparently waterbome particles could initiate epizootic levels of perkinsiasis during late summer in a deployment of experimental oysters. Any successful management strategy must consider that epizootics can be initiated by two mechanisms: 1) rapid prolifer- ation and spread of enzootic infections within a local subpopula- tion. or 2) advection of waterbome infective stages of the parasite from other infected subpopulations of oysters. The spatial scale on which the latter mechanism operates is unknown. The invasion of P, mannus into low salinity seed oyster areas was a turning point in the host-parasite relationship within the Chesapeake Bay system (Andrews and Ray 1988). Once heavy levels off. mannus infection occurred in seed oysters, the patho- gen was spread rapidly throughout the Chesapeake Bay system by the practice of planting private and public seed beds. Based on recent surveys (Burreson 1989. Krantz 1993) it appears that P. mannus is established throughout Chesapeake Bay. with the high- est intensity of infections on some of the best oyster-growing areas. Annual, seasonal, and spatial variations in salinity within the Bay suppress the impact of the disease on some host popula- tions. However, once a specific population becomes infected with P. marinus it may be expected to experience epizootic losses whenever environmental conditions favor the parasite. Therefore, because P. marinus never can be eradicated from enzootic areas, management should focus on activities that will reduce the losses of oysters in the expectation of epizootic levels of perkinsiasis (Andrews and Ray 1988). Temperature apparently is a regulating factor for P. marinus prevalence and infection intensity. Warmer water temperatures in the Gulf of Mexico increased the intensity and duration of perkin- siasis even at lower salinities (Andrews and Ray 1988). Recent experience in Maryland has supported the idea that cold winters and high rainfall are mitigating factors. The prevalence and inten- sity of P. marinus. which had reached epizootic levels throughout Chesapeake Bay by 1992, decreased considerably in 1994 (at least in Maryland waters) after two successive wet. cold winters. Im- proved growth and survival during 1994 led to geographic expan- sion of the commercial harvest during the 1994—1995 season, along with increased landings (Krantz 1995). Alternate molluscan hosts in the natural environment can be important as reservoirs for P. marinus (Perkins 1988, Andrews and Ray 1988). Andrews (1954) found putative P. mannus cells in a majority of the potential molluscan host species that he sampled in the Chesapeake Bay system. Many of these species are natural inhabitants of Chesapeake Bay oyster bars, and some develop dense populations in the muddy bottom immediately adjacent to natural oyster bars. When an oyster-planting site has been depop- ulated (e.g., by disease), the remaining shell base rapidly becomes heavily colonized by other molluscan species. Mollusks parasit- ized by P. marinus-Vi\^c organisms appear to shed cells from their intestinal tract throughout the summer and could be a significant source of infectious material in the immediate vicinity of any managed oyster bar. The question remains whether all of these molluscan parasites are in fact P. marinus or closely related species that do not para- sitize oysters. White et al. (1987) found a gastropod, Boonea impressa. that was capable of transmitting P. marinus from one oyster to another or from populations of the ectoparasitic snail, a natural inhabitant of oyster bars. Antisera against P . marinus cells cultured in vitro recently were used to assess the serological sim- ilarity of the P. marinus parasites found in other molluscan spe- cies: several Chesapeake Bay mollusk species contained organ- isms which cross-reacted strongly with antiserum prepared with P. marinus cultured from oysters (Dungan and Roberson 1993). MANAGEMENT CONCEPTS In a discussion of management strategies to control levels of mortality caused by P. marinus. Andrews and Ray ( 1988) focused their recommendations on the importance of planting disease-free oysters on cultivated barren bottom or on natural oyster bars from which all previous generations of diseased oysters had been re- moved. At the time of their recommendations, a few sanctuaries of seed oysters free of P. marinus infections existed in Chesapeake Bay. By the early 1990s, however, virtually all natural oyster populations contained some oysters with detectable levels of P. marinus disease. These authors concluded that "... control [of perkinsiasis epizootics) depends upon the return of normal rainfall and low estuarine salinities to suppress the disease which remains active during summer at levels of 12 ppt ..." .^ndre\^s and Ray ( 1988) outlined management procedures for controlling P. marinus. with the following recommendations: "1. Transplant only disease-free stocks. Even low preva- lence of P. marinus will accelerate mortality. "2. Select growing beds isolated at least 0.4 kilometers from any other bed with infected oysters. "3. Early harvest (2-3 inches) followed by tallowing of beds limits mortality and distribution of the disease. "4. Monitor beds of growing eastern oysters for P. mari- nus. All planted beds should be examined . . . each late summer, or early fall for dead animals to deter- mine if harvesting or another year of culture is desir- able or necessary." These recommendations were focused on controlled cultivation of private, leased beds where the removal of diseased oysters and the planting of only disease-free oysters could limit mortalities caused by P. marinus. Public oyster bars, however, are never completely harvested because of regulatory minimum size limits and gear inefficiencies, and because natural recruitment maintains a population of small oysters that serves as a reservoir for the parasite. Thus, control of P nuumus is more difficult on public beds than on private beds. APPLIED MANAGEMENT PRACTICES AND THEIR Ol'TCOMES Natural oyster bars in the Maryland portion of the Chesapeake Bay are the mainstay of a public fishery, with legally defined areas where only specific types of gear (hand tongs, patent tongs, oyster dredges, or hand collection by divers) can be used to harvest oysters. Hand tongs are extremely inefficient, and even under high harvest pressure, a significant percentage of adult oysters remains on the bottom following the prescribed harvest season. Patent Management Protection of Oyster Fisheries 169 tongs are slightly more efficient at collecting oysters, but still a large percentage of the population remains. Power dredging and scuba divers are much more efficient on specific bottom types. However, any legal harvest operation will leave pockets of In- fected oysters on the natural bars. Smith and Jordan (1992) esti- mated that 53% of harvestable oysters (S76 mm) were taken from harvested oyster bars (all gears combined) in Maryland during 1990-1991. For many years, oyster managers have depended on movement of oysters from established seed areas to growing areas. In earlier decades, the main purposes of this practice were I) to relieve overcrowding in areas of very high recruitment, where oysters did not grow well because of competition for food and space, and 2) to distribute the harvest geographically according to the locations of private growers or public fisheries. When P . inanntis and, to some extent, Haplosporiiliiim nclsoni (MSX) became epizootic in the Chesapeake, this practice was adapted to increase survival of stocks to harvestable size by concentrating growing areas in lower salinity reaches of the Bay and tributaries. Maryland instituted a new practice of producing seed oysters for the public fishery by planting dredged fossil shell on hard bottom with few if any ex- isting natural oysters. The seed plantings now are made several miles from heavily infected natural oyster populations. Several of the most recent plantings have been made on the western side of Kedges Straits (Fig. I): production of seed oysters has been very successful. Movement of these seed oysters, even though lightly infected with P. marinus and H. nelsoiu. to areas of low salinity in the upper reaches of the mainstem of the Bay and other subeslu- aries has been successful in producing harvestable oysters (Krantz 1995). In the Chester River this strategy has maintained a public fishery even during a series of summer drought conditions that induced epizootic levels of P . nuiniuis over all of the growing grounds in Maryland (Fig. 2). Figure I. Northern Chesapeake Bay showing the location of the prin- cipal current ( 1995) Maryland seed oyster production area. Figure 2, P. marinus prevalence (%) in Maryland oyster populations in the fall of 1992. based on thioglycollate incubations of hemolymph samples from 30 oysters at each of 43 monitoring sites. Concerns about spreading pathogens through movement of in- fected seed stocks, and strong interest in relieving growing areas from the pressure of P moriiuis. led the Maryland Oyster Round- table (Maryland DNR 1993) to establish six Oyster Recovery Ar- eas in Chesapeake tributaries (Fig. 3). Zones were designated in each of these tributaries where no diseased seed oysters (defined as oysters with zero prevalence of P . marinus and H. nelsoni as determined on a random sample of seed by standard diagnostic methods) were to be planted. Because production of disease-free seed oysters in the natural waters of Chesapeake Bay cannot be assured, experimental quantities of seed oysters are being pro- duced in hatcheries, with a long-range goal of supporting seed requirements for the disease-free zones, as well as for private growers. Ambient water, however, is used to grow the newly set spat in all of Maryland's oyster hatcheries. Therefore, contamina- tion of hatchery-produced seed by waterborne P. marinus could occur during the postsettlement and early growth periods. Little is known about P. marinus infections in very young oysters or lar- vae, but if hatchery seed Is grown in water where P marinus is enzootic, it is likely that they will become Infected within the first year (Krantz 1993). Selecting oyster-growing beds that are isolated from other beds is difficult where natural oyster bars are the basis of a public fishery, as in Maryland. Although oyster "bars'" In parts of the Atlantic and Gulf coasts are discrete features of the estuarine sea- scape, in large areas of Chesapeake Bay there are nearly contin- uous populations of oysters In much of the habitat, albeit with large variations in the density of oysters. Recent studies on water- borne particles in the Maryland portion of the Chesapeake Bay Indicate that putative P marinus particles are present in the water column from April through October, with peak concentrations in 170 Krantz and Jordan River \^^ Severn River l»"y Figure 3. Maryland Oyster Recovery Area tributaries with approxi- mate zone boundaries. See text for zone delinitions. June (Roberson et al. 1993). Burrcson and Andrews ( 19X8) found that spatial isolation ot groups of oysters did not protect them from P. marinus during drought conditions in 1985-1987. This was a period of geographical expansion of P. nuiriinis into uninfected populations in the low salinity sanctuaries in Maryland. Evidence, albeit indirect, has accumulated that walerborne concentrations of P . inannus may be more important in establishing epizootic levels of the disease than proximity of infected natural oysters to the newly planted seed. Early, exhaustive harvest of planted beds and allowing the beds to remain fallow between plantings could not be an option for a public fishery without major changes in law and management practice. For example, the Maryland 3" (76-mm) cull law has been in place since the early part of the 20th Century and has been thought to protect the spawning potential of the stock. The prob- lem is that perkinsiasis, alone or in concert with MSX, crops a significant percentage of a given year class before it reaches 76 mm. Lowering the legal size to. say. IVi (64 mm) would allow increased exploitation of younger year classes while increasing the abundance, if not the biomass, of the harvest. Exploitation of the smaller oysters, however, could decrease the recruitment potential of the population. Further, it has been suggested that minimum size limits cause selective pressure for smaller size at maturity (Allen and Bushek 1994); if so, then decreasing the cull size could exacerbate this effect. A decrease in the cull size also could in- teract with the host-parasite relationship by providing fewer op- portunities for oysters to survive P . marinus infections, reproduce, and thereby expand any genetic resistance, should it exist, in the population. A "slot limit" (e.g., 2Vi-A" or 64—102 mm), as pro- posed for evaluation in the Maryland Oyster Recovery Action Plan (Maryland DNR 1993). would protect some larger, more fecund, and perhaps resistant oysters. It Is questionable, however, whether there are enough oysters >102 mm remaining in Chesapeake pop- ulations to mitigate the negative effects of a smaller harvest size. A smaller size limit also could lower the retail value of shucked oysters to a point where it would make the fishery economically unattractive to both processors and watermen. Virginia imple- mented a reduction in size limit to 2'/:" for the James River seed area in 1991. This action resulted in the taking of increased amounts of James River seed. In the following years, harvests and recruitment were greatly depressed, and Virginia now lacks natu- ral seed oysters to implement improved management on any sig- nificant scale MONITORING OF P. MARINUS Andrews and Ray ( 1988) suggested monitoring oysters 2 years of age or older by the thioglycollate culture method in late summer and early fall. They suggested that the beds be concurrently ex- amined for mortality. Population mortality rates and disease prev- alence then could predict whether planted oysters could survive and grow another season or had developed high intensity infec- tions of P. miiriiuis that would predispose them to high mortality during the coming months. If oysters were found to be heavily infected, then they should be harvested regardless of their size or meat quality, if prevalences and mortality were low, indicating an enzootic P. maniiHs infestation, the grower could risk leaving the oysters in place until the spring and early summer of the following year. After 2 years of exposure to P. marinus. however, summer mortalities can be significant, and any population found to be enzootic for P . marinus in the fall should be monitored early in the following summer. Maryland implemented a comprehensive disease-monitoring program for natural oy.ster bars in 1987. It was designed to doc- ument and follow recently established P. marinus infections, as well as the impacts of epizootic P . marinus in populations where the parasite had been established for several years. This monitor- ing program was instrumental in determining the geographic ex- pansion of perkinsiasis in Maryland waters and has delineated areas within subestuaries where P. marinus remained at low en- zootic levels and induced only minimal levels of mortality in nat- ural oyster populations. The monitoring program has continued to follow seed oysters planted in both enzootic and epizootic portions of Maryland waters and documented site-specific changes in prev- alence and intensity over an H-ycar period (Krantz 1989a. 1990. 1991 . 1993. 199,'i). The monitoring program was modified in 1990 to produce consistent annual observations on selected oyster bars regardless of their disease status. Information from the fall field surveys on population structure, size frequency, mortality, and spatfall is now collected from 64 stations distributed throughout the Maryland portion of the Bay. A subset of 43 of these stations is monitored for P . marinus and H . nelsoni prevalence and inten- sity (Fig. 4). These stations were originally part of a long term monitoring effort to document oyster recruitment dynamics on natural bars in the Maryland portion of the Bay. The monitoring program is linked to a geographical information system (Smith and Jordan 1992) that includes several attributes of oyster habitat such as water quality, bathymetry, and substrate type, in addition to the population and disease data. .Analysis and modeling studies are in progress to better understand the dynamics of P. marinus in oyster populations and to evaluate the effectiveness of management strat- egies, both tried and proposed. In areas known or suspected to harbor P . marinus (virtually the entire Atlantic and Gulf coasts), seed oysters to be transplanted should be monitored for diseases before thev are moved. The seed Management Protection of Oyster Fisheries \ ^^'■-^ ♦''s t< *^ "/'nI J J 1. ■ ■ 1 1 ./ \. i ..>' ?itr '.^* ^r ^t '^^t h 'V 1 ■I &''ii'iS^ ^■:.'p |^.|/.!--^ Figure 4. Maryland Kail ()\ster Sur\e> population and disease- monitoring sites. Circles: population samples (size, mortality, spatfall, etc. I only: crosses: population and disease samples. area should he evaluated during the tall prior to seed movement for prevalence and Intensity of diseases. After planting, disease status and mortality should be monitored in the fall of the secc)nd and third growing seasons. NEW DIRECTIONS IN MANAGEMENT Oyster management in Chesapeake Bay has taken on a posture of avoidance of areas where the disease is epizootic and has fo- cused plantings of seed oysters in areas of low P. maniuis pres- sure, the equivalent of low salinity sanctuaries described by An- drews and Ray ( 1988). The collective experience of previous sur- veys for disease prevalence, oyster mortality, and response of planted seed oysters to documented levels of fall disease preva- lence and intensity permitted the delineation of management areas in Maryland (Fig. 5). The four areas were based on the combined effects of P. marinus and H. nelsoni during the past two decades. Each year's management recommendation for the areas is tem- pered by the results of the previous Fall Survey, which may detect diminution or increases in prevalence and intensity of the diseases in the areas. The recent report of the 1993 and 1994 Maryland surveys (Krantz 1995) is an example of how monitoring technol- ogy can assist the management agency in selecting areas that will optimize expenditure of funds for the placement of seed oysters, shell for the collection of natural spat and oyster bar rehabilitation, and the development of seed-producing areas. Although oyster population and disease monitoring in Mary- land is reasonably comprehensive, information on salinity, tem- perature, and water quality patterns throughout the year could help management to anticipate disease and recruitment conditions. A geographical information system (developed for managing oyster- Figure 5. Maryland Chesapeake Bay oyster management areas (1992) based on potential disease impacts (adapted from Krantz 1993). The areas are numbered according to potential disease impacts, with .4rea I representing the safest growing areas and .Area 4 representing areas most likely to be affected by both P. marinus and H. nelsoni. .Areas 2 and 3 represent increasing risks of parasite-induced mortality, pri- marily from /'. marinus, although //. nelsoni can affect these areas during drought periods. monitoring data) is being used to develop spatial interpolations of the extensive Chesapeake Bay water quality monitoring database (Heasly et al. 1989). in combination with modeling studies of population and disease dynamics. Eventually, these efforts should provide some ability to predict oyster infection prevalence and intensity, mortality, and population structure from knowledge of seasonal and interannual environmental conditions (Hoffman et al. 19951. Maryland, like most mid-Atlantic coastal areas, relics on nat- urally set seed for planting oyster bars that received light or no spat set. yet are in very productive growing areas. These areas have been identified by annual fall surveys in Maryland waters that began m the mid-193()s. Collective knowledge and memory of the performance of plantings at specific sites are passed from one generation of oyster management biologists to another through practical field exposure. Seed production areas in the state require a comprehensive knowledge of the past history of the natural oys- ter bars in that area, changes in the bottom type, the potential impact of storm events during the winter and early spring, envi- ronmental water quality conditions during the summer setting area, and most important, the past history of natural spat set. The investigation of natural spat set is one of the pursuits of the Fall Survey, and historical information has been used to locate areas that have the most probable chance of acquiring a natural spat set. Detailed information on natural spat set on the oyster bars has been described by Krantz and Meritt ( 1976) and is currently being up- dated by the senior author for publication. During the past three decades, seed areas in the Maryland portion of the Bay that have produced a cost-effective quantity of spat on planted shell have 172 Krantz and Jordan changed dramatically. Historically, the managenient agency had the choice of producing seed in St. Marys River, upper Tangier Sound, Honga River, Little Choptank River, Harris Creek. Broad Creek, and Eastern Bay (Fig. 6). Naturally set seed oysters first showed a decrease in numbers and frequency of set in St. Marys River, Eastern Bay, and portions of Tangier Sound. Durmg the past decade, seed production has shown a radical decline in Harris Creek and Broad Creek and a diminution in the Little Choptank and Honga Rivers. As this decrease in spat set occurred, a con- current increase in spat set on sparsely populated bottom along the Chesapeake Bay side of St. Marys County and in lower Tangier Sound was noted. As a result, state-managed seed areas now have been moved to areas where the highest spat set can be obtained. Substrate for seed production areas is critical to continuous production of seed to be transported. Shell is collected from all Maryland processing and shuckmg plants, to be planted back in the Bay. However, as harvest has decreased, the quantity of freshly shucked shell has become inadequate for seed production. Unfortunately, fresh shell is dispersed throughout the entire Ches- apeake Bay. and the cost of hauling all of the shells to the seed beds is excessive. Therefore, fresh shell from packing houses is now used to rehabilitate oyster bars located near packing houses, or to add to the planting of dredged fossil shell on seed beds. Dredged fossil shell is obtained in the upper Chesapeake Bay through a contract with a large dredging company that has the capability of digging 35-40 feet below the Bay surface and esca- lating oyster shell to the surface. The shell is then graded, washed, and transported by barges and tug boats to areas selected by the management agency to be planted. Annual planting of dredged shell on the best seed areas usually produces 500-2.500 spat per bushel of dredged shell material at a given location. Spat set on natural oyster bars ad|acent to these areas would range between 50 and 250 spat per bushel of shell during the same season. The J J ' fA'^stern fBaif ^^'^^^fptrociqf Creeks ''; v4'\ Little Clidpfank River (*' Figure 6. Areas used historically for natural seed oyster production in Maryland's Chesapeake Bay. seasonal periodicity of planting the dredged shell has been a point of great concern. Maryland has accumulated records of periodicity of spat fall on natural oyster bars, planted shell, and experimental spat collectors for over three decades. These data show 90% of the annual recruitment in any given year occurs between the second week of June and the end of July. Shells planted during this lime will usually receive an adequate concentration of spat to be used as seed the following spring. By law. seed areas may not be located on productive natural bars In Maryland waters. Therefore, considerable effort must be expended to find firm bottom capable of supporting dredged shell. Frequently, natural oyster bars that have been heavily harvested or devastated by disease become covered by a layer of sediment 5-10 cm in depth, but they can be used as seed areas. The planting of dredged shells to a depth of 10-15 cm on top of this substrate provides a base for oyster attachment, but at the expense of losing some of the shells in the reconditioned substrate. Records of shell placement to establish new seed areas revealed that only 'A to '/: of the planted shell can be recovered with spat on it. A high percentage of shell settles into the soft surface and becomes anaer- obic and unsuitable for the attachment of spat. Two decades ago, there were areas in the Chesapeake Bay in Maryland and Virginia that had relatively high spat fall and no disease in the natural oyster populations. Since 19S0. this situation has changed. Andrews and Ray ( 1988) pointed out that the ability to manage around perkinsiasis was greatly altered when all of the seed area became infected. Even the low salinity sanctuaries in the Chesapeake Bay now have occasional and temporary excursions of P. mariinis in the natural populations that establish reservoirs to infect seed oysters. Maryland management agencies have found by trial and error that new seed areas can be established annually If the areas are carefully chosen. Spat settlement in Maryland waters is between June and the end of July. By fall of the first year, some of these seed areas have been demonstrated to be free of detectable levels of P. marimis. In certain years, some seed areas will acquire low prevalences of very light infections (up to iO^c). These seed can be moved the following spring, prior to the biologically active season for P. marimis development, and placed on oyster bars for growth. Current policy, however, does not permit planting of oysters with any detectable level of P . marimis Infection in Zones A or B of Oyster Recovery Areas (Fig. 3). Occasionally, spat set on a selected seed area is light and the number of seed is below the level that is economically attractive for movement (350-500 spat per bushel). Such seed beds will be left for another biological season in which more spat will attach to the shell. In these situations, it has been the experience in Mary- land waters that P. marimis will increase in prevalence and inten- sity during the second biological season. Often, seed that remained on the seed areas for two seasons had a prevalence of lOOVr . with infections at lethal severities in 30% of the animals. In other cases, seed can remain on the dredged shell for two biological seasons without contracting detectable levels of P. marimis. It is rare, however, to find a seed area that is free of disease, because all oyster bars surrounding productive seed areas in Maryland have high prevalence and intensity of P. marimis infection, in sum- mary, the management agency has encountered four distinct classes of seed beds, which vary both spatially and from year to year: 1. Seed areas with high prevalence and high intensity of P. manniis infections. Management Protection of Oyster Fisheries 173 2. Seed areas with high prevalence, but low intensity of P . mar inns. 3. Seed areas with low prevalence ot P nutniuis. but infected animals have severe infections. 4. Occasionally seed areas will have the most desirable disease status — low prevalence of P . nuirimis with low severilv of infection. The general outcome of moving the above four categories of seed can be predicted based on general observations of cumulative mortality of oysters caused by epizootic levels of P. marinus in Maryland waters (Table I). Table I summarizes field observa- tions, along with experimental exposure of oysters in bags and trays at various sites throughout Maryland during the last decade. In some cases, all of the oysters would be destroyed by a combi- nation of//, nelsoni and P. marinus in two seasons, leaving onlv a few trays free of MSX. from which the data on P marinus were taken. However, a voluminous amount of data on the relationship between prevalence and intensity of pcrkinsiasis and resultant mortality have been accumulated by the annual Fall Survcss (Krantz 1991. 1993. 199.^^; Smith and Jordan 1992). The outcome of transplanting seed oysters with a high preva- lence and intensity of P. marinus infections will be moderated by both the environment and the disease pressure of the area in which they are planted. Seed oysters with a high prevalence and high disease intensity routinely develop a prevalence in excess of 65% during the first year of life. Population mortality during the first and second biological seasons will reach levels shown in Table 1 . By the third biological growing season, cumulative mortality of seed oysters with these characteristics may reach 90-95"^^ and very few of the animals will enter the annual harvest. Transplantation of seed oysters with a high prevalence and low severity of P. marinus usually requires 2 years at the new site before the population reaches epizootic levels with 20-30% severe infections. By the time the third biological season is completed, population mortality may range from 50 to 75%. Planting seed oysters with a low prevalence but with a high severity of infection usually follows a similar track. By the end of the second year, the population reaches epizootic status and begins experiencing high mortality. During the third biological season cumulative mortalitv may exceed 50-60%. Planting of seed oysters with low prevalence and infection in low salinity environments will permit the popu- lation to grow at normal rates for the area until the third biological season, in which time the population may reach epizootic levels, with losses of 10-30% of the animals by the time they reach harvest size. Seed oysters with low prevalence and severity of P. marinus infection may show no significant mortality when transplanted into a low salinity area typical of the sanctuary zones described by TABLE 1. Expected cumulative mortality { '7c ) caused by oyster diseases at epizootic levels In Maryland waters, based on a s> nthesis of several years of observations from natural populations and experimental deployments by the senior author. \ ear Parasite 1 2 3 4 P. marinus H. nelsoni 27 46 58 77 81 89 92 94 Andrews and Ray (1988). This concept is the basis for Maryland's present oyster management program, in which seed oysters are taken from setting areas that traditionally have high prevalence of P. marinus. This seed is planted in low salinity sanctuaries. The Chester River (Fig. 3) is one of the areas with chronic P. marinus infections, but prevalence and intensity are low enough so that oysters will grow to market size in a 3- to 4-year period with a loss of approximately 30% of the transplanted seed. Maryland has been able to maintain a "put-and-take" fishery in the Chester River and the upper Chesapeake Bay mainstem through the movement of lightly infected seed oysters. Table 2 shows the results of main- taining a "put-and-take" fishery while the harvest from natural bars in Maryland's waters dropped from 1 .5 million bushels in the 1985-1986 season to 65.000 bushels in the 1993-1994 season. Less than 25.000 bushels of seed oysters were planted in the Chester River in 1986. with a harvest about equal to this input. At this point, the decision was made to stop planting oysters in the lower Bay. where high disease prevalence was killing the oysters before they reached market size. The seed that would have been planted in these areas was then relocated to the Chester River where a fishery resulted that by 1993 accounted for over 50%- of Maryland's harvest. In earlier years, the Chester River could hardly be considered an area to support significant sustainable harvests because of very low annual recruitment. From 1985 to 1995. the cost of collecting and hauling the seed oysters from the lower Bay seed areas has risen from $1.00 to $1.30 per bushel for the contracted labor and vessel. In addition, there is a highly variable cost for the amount of shell that is placed on the seed area from which the seed is taken. In some cases, as much as 50% of the shell can be recovered and moved, but in most cases only 30% of the planted shell will have seed on it. Dredged shell costs have been fairly stable during the past decade, ranging from 18.5 cents per bushel to the present cost of 26 cents per bushel. The "put-and-take" fishery being maintained in the Chester River has been characterized as a subsidized fishery, since the Department of Natural Resources has expended more than the severance tax (45 cents per bushel) gained from the packer who bought the oysters. For instance, in 1 986. 24.695 bushels planted at a cost of $29. 140 produced 50,679 bushels of harvested oysters (Table 2). The planting cost of $0.57 per bushel harvested was not recovered from the $0.45 per bushel severance tax. although other indirect benefits such as employment, income taxes, and economic TABLE 2. Oyster production in the Chester River < Maryland I in response to planting; natural seed oysters. The oysters were harvested 2-3 years after planting. Units are Maryland bushels. Year Seed Oysters Planted Bushels Harvested 1986 1987 1988 198y 1990 1991 1992 1993 1994 24.695* 55.465 155,530 114.675 37.770 83.590* 161,684 62.089 74.431 24.738 20.597 50.679 54.029 60.468 55.123 53.803 51.271 Not available Supplemented by natural spat fall 174 Krantz and Jordan multipliers would have to be considered in a realistic economic analysis. If a state agency is considering management of oysters under conditions influenced by epizootic perkinsiasis, it must be pre- pared to subsidize the immediate costs of maintaining the fishery. One of the most important aspects of management of an oyster fishery during a P. marinus epizootic is the establishment of new seed areas each year, to produce a single year class of seed that will have a low level of infection. If dredged shell are planted on top of a previous year class, a percentage of the older year class with higher prevalence and more intense infections would be mixed with the new year class being produced by the seed area. However, the re-establishment of the new seed areas each year utilizes a greater percentage of shell resources as base material and contributes to the increased cost of the seed. The selection of the geographical location is an empirical and qualitative type of deci- sion. All of the past experiences with planting and production of seed oysters, as well as the most recent monitoring data on the disease prevalence for seed and oysters on natural bars adjacent to the seed area, must be considered. Spat fall is highly erratic and virtually unpredictable from season to season in the Maryland portion of the Chesapeake Bay (Krantz and Meritt 1976. Krantz 1992. Dekshenieks et al. 1993). At present, the Maryland man- agement agency selects two or three new seed production sites each year, in an attempt to produce the greatest concentrations of seed with the lowest prevalence and intensity of P . marinus. In addition to the use of a disease-monitoring program and production of uninfected or lightly infected seed, some states use a quarantine zone {P. iminnus-fTec) for producing natural seed oysters. No diseased oysters are introduced into the seed area and seed could remain in that area and not be transferred into the zone where P. marinus was present. This situation exists in Delaware Bay and some portions of Long Island Sound. Unfortunately, many areas in Delaware Bay that are free of P. marinus have low spat set, at concentrations that are not economically attractive to be moved. A quarantine concept could also be used to protect New England stocks from contamination by seed oysters or brood stocks from areas where P. marinus is epizootic or enzootic. The Atlantic States Marine Fisheries Commission initiated a shellfish transport management plan in 1989 that could be implemented to assist in establishing quarantine procedures to protect oyster- growing areas from importation of disease (Krantz 1989b). At the present time, the plan has been approved by Atlantic States Marine Fisheries Commission, but funding has not been appropriated to implement the committee to oversee and coordinate management among the New England states. This program would have a great impact on prevention of spreading P. marinus into new areas. Low salinity culture of oysters is a possible alternative. Essen- tially, a program such as described above for Maryland could be implemented by the private sector. State and private sector orga- nizations, working cooperatively, could establish hatcheries and seed areas in low salinity environments where the impact of P. marinus is minimal. The biological cost of this approach is usually a slower growth rate and poor, or at least unpredictable, meat quality. Low salinity culture sites must be very carefully selected and observations on small quantities of oyster seed and adults should be made over a period of years prior to the investment of large sums of capital. This strategy would allow management and invt.^tors in the low salinity mariculture system a chance to expe- rience the natural diversity of responses of growth, disease fluc- tuations, spat set, and survival in the low salinity areas. Low salinity areas that inhibit P. marinus are usually those which are not optima! for the growth of oysters. Two to three additional years of growth are required in low salinity environments in Mary- land waters for planted oysters to reach market size, compared to oysters grown in areas that are now heavily infected with P. mari- nus. Technology for use of disease-free brood stock and oyster pro- duction in hatcheries is well known, but this approach has not been cost effective in the past (Krantz 1982). Closed-system hatcheries have the capability of maintaming brood stock and seed oysters that have not been exposed to disease. The costs of operating closed-system seed production has not been evaluated at this time. The greatest shortcoming to the hatchery approach is still the en- vironmental setting in which the disease-free seed could be grown at their optimum. The low salinity environments with character- istics of the sanctuaries proposed by Andrews and Ray ( 1988). or the present growing grounds used in the Maryland portion of the Ches.ipeake Bay, arc areas where oyster growth rate and meat production are not very good. Therefore, the economic return from an expensive hatchery seed oyster may not be realized when placed in the natural environment. Attempts to grow seed oysters to market size in hatcheries have been conducted in the past, and none of these have been shown to be cost effective. Oyster geneticists continually express interest in developing oyster strains that are resistant to perkinsiasis. Andrews and Ray (1988) speculated on the possibility of finding strains of oysters that were resistant to perkinsiasis. In their studies, they found there were always oysters that survived epizootics. However, after studying 40 years of natural selection in the Gulf of Mexico, Ray responded to a question at the 1993 National Shellfisheries Asso- ciation Annual Meeting in Charleston. SC. that he has never found a population of oysters in the Gulf of Mexico to become resistant and show a diminution in the prevalence of P. marinus and sub- sequent mortality to perkinsiasis. Andrews reported on numerous occasions to have obtained stocks of oysters from Virginia waters that acquired heavy infections of P. marinus. but that did not have much resulting mortality (e.g.. Andrews 1965. 1967). Offspring from these strains appeared to have better survival than uninfected seed oysters that were placed in trays for comparison. However. Andrews" experiments always were compromised by the interac- tion of H. nelsoni that cropped a large percentage of his test animals. Interest in development of P. marinus-resislaiw. strains has recently intensified. For example. Bushek et al. (1994) eval- uated the host-parasite interaction for strains from a range of East Coast and Gulf Coast locations and determined that oysters and parasites, respectively, had heritable variations in resistance and virulence. Recently, a subpopulation of oysters from the Nanti- coke River in Maryland was observed to have survived to large size (>102 mm) in an area that had experienced P. marinus epi- zootics for several years. Selective breeding and evaluation of Fl and F2 generations for P. marinus resistance are in progress, in a Maryland Department of Natural Resources and National Marine Fisheries Service cooperative investigation. Despite the hopes, frustrations, and in some cases, moderate successes of attempting to manage around perkinsiasis. nature still has the upper hand. For example, after several successive years of record low harvests in Maryland in the late 1980s and early 1990s, the harvest more than doubled between the 1993-1994 (-70.000 bushels) and 1994-1995 season (—150,000 bushels by preliminary Management Protection of Oyster Fisheries 175 estimates). This reversal followed two consecutive years (1993 and 1994) of high spring rainfall and runoff. Salinities in the northern Chesapeake Bay and tributaries dropped to the point where the Maryland Bay-wide average of P . marinus prevalence decreased from 84 to 53%, and H. ncLsoni virtually disappeared after causing a record epizootic in 1992 (Krantz 1995). Harvests not only increased, but several areas produced harvests that had not been productive lor a decade or more. CONCLUSIONS Clearly, there remains great potential for natural recruitment and recovery of oyster populations m the northern Chesapeake Bay, despite the depredations of the parasites and the often-voiced concerns about overharvesting a depleted resource. This potential probably exists in other oyster-growing areas, whether assisted by rational management strategies or not. Harvesters, packers, mar- kets, and managers, however, cannot thrive (or survive?) in the face of the uncertainty of waiting for the next freshet. The prob- lem, then, is to establish stable harvests that are minimally subject to natural variations in climate, recruitment, and parasitic infec- tions. Maryland's dominant oyster management strategy of recent years — establishing new seed beds and transplanting to low salin- ity growing areas — has resulted only in slowing the decline in harvests. The cost effectiveness of this technique has been mar- ginal, at best. It remains to be seen whether the state's new oyster recovery strategy of establishing quarantined areas and sanctuar- ies, and encouraging and enhancing hatchery seed production, research and monitoring, will succeed. Experiences in the Ches- apeake, the Gulf of Mexico, Long Island Sound, and probably other East Coast oyster-growing areas indicate that management of oyster populations infected with P marinus should include the following elements: 1 . Optimizing natural recruitment by developing and maintain- ing clean seed beds free of infected older oysters: 2. Preventing movement of infected stocks into growing ar- eas— in some cases growing areas may need to be depop- ulated and left fallow for a time prior to planting; 3. Enhancing the capability of local hatcheries to supplement natural recruitment and to provide uninfected seed oysters and larvae for mariculture operations: 4. Careful monitoring of oyster populations and seed stocks, along with important environmental information such as temperature, salinity, dissolved oxygen, and turbidity; 5. Directed research, particularly in the areas of genetic resis- tance to P. mariiuts and improved brood stocks, more ef- ficient diagnostic methods, and the ecological dynamics of host-parasite-environmental interactions . Several major contributions have been made very recently to our knowledge of P. marinus and our ability to detect and exper- iment with the parasite. Genetic studies and selective oyster breed- ing are promising to produce a more resistant oyster, or at least to generate a better understanding of the problem. Improved moni- toring programs are yielding excellent information on oyster pop- ulation and disease dynamics on both system-wide and local scales. Modeling studies are applying this new knowledge to begin to predict outcomes of management strategies in the face of natural variability. New tools, new knowledge, and new skills are making P . marinus a more tractable problem and may soon result in prac- tical approaches to better management of oyster stocks. What is perhaps most hopeful in Maryland is that the Oyster Recovery Action Plan has laid the groundwork for an unprece- dented partnership between the public fishery, private growers, managers, scientists, and environmentalists, with the goal of re- storing both the ecological and economic benefits of oyster pop- ulations. Long term cooperation between these groups could en- sure that scientific advances are applied responsibly and that con- cerns about overharvesting, habitat loss, and conflicts between the public and private fisheries are addressed jointly, rather than com- petitively. These groups are sharing information and working to- gether to plan and implement moderate-scale oyster enhancement projects. New developments in research, monitoring, manage- ment, and policy, however promising, certainly will not make P . marinus (or H. nelsoni) go away, but they may, given enough time, result in larger and healthier oyster populations. LITERATURE CITED Allen, S. K. & D. Bushek. 1994. Resistance of Crassostrea virginica races to Perkinsus marinus isolates: a foundation for breeding and management. Final report to U.S. Dept. of Commerce, NOAA National Marine Fish- eries Service, Northeast Region, Grant No. NA26FL0381. Andrews. J. D. 1954. Notes on fungus parasites of bivalve mollusks in Chesapeake Bay. Proc. Nail. Shellfish. .Assoc. 45:157-163. Andrews, J. D. 1965. Infection experiments in nature with Dermocysiid- mm marinum in Chesapeake Bay. Chesapeake Set. 6:60-67. Andrews, J. D. 1967. Interaction of two diseases of oysters in natural waters. Proc. Natl. Shellfish. A.s.soc. 57-38-49. Andrews, J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marinus and its effects on the oyster industry. Amer. Fish. Soc. Spec. Publ. 18:47-63. Andrews, J. D. & S. M, Ray. 1988. Management strategies to control the disease caused by Perkinsus marinus. .Amer. Fish. Soc. Spec. Puhl. 18:257-264. Burreson, E. M. 1989. Prevalence of the major oyster diseases in Virginia waters — 1988: Summary of the annual monitoring program. Vir. Inst. Mar. Sci. 6 pp. -f tables. Burreson, E. M. &J. D.Andrews. 1988. Unusual intensification of Ches- apeake Bay oyster diseases during recent drought conditions. Proc. Oceans '85 Conf: pp. 799-802. Bushek, D., S. E. Ford & S, K. Allen. 1994. Evaluation of methods using Ray's tluid thioglycollate medium for diagnosis of Perkinsus marinus infection in the eastern oyster. Crassostrea virginica. Annu. Rev. Fish Dis. 4:201-217. Craig, A, E. N. Powell, R. R. Fay& J. M. Brooks. 1989. Distnbution of Perknisus marinus in Gulf Coast oyster populations. Estuaries 12:1 82- 191. Dekshenieks, M. M., E. E. Hofmann & E. N. Powell. 1993. Environ- mental effects on the growth and development of eastern oyster, Cras- sostrea virginica (Gmelin, 1791), larvae: a modeling study. J. Shell- fish. Res. 12:241-254. Dungan, C. F. & B. S. Roberson. 199.V Binding specificities of mono- and polyclonal antibodies to the protozoan oyster pathogen Perkinsus marinus. Dis. Aquat. Org. 15:9-22. Fisher, W. S., J. D. Gauthier & J. T. Winstead. 1992. Infection intensity oi Perkinsus marinus disease in Crassostrea virginica (Gmelin. 17911 from the Gulf of Mexico maintained under different laboratory condi- tions. J. Shellfish Res. 1 1(2):363-369. 176 Krantz and Jordan Heasly. P.. S. Pultz & R. Batiuk. 1984. Chesapeake Bay Basin Monitor- ing Program Atlas. Vol I. Water Quality and Other Physiochenilcal Monitoring Programs. U.S. Environmental Protection Agency CBP/ TRS 34/89, Annapolis. Maryland. Hoffman. E. E., E. N. Powell. J. M. Klinck & G. Saunders. 1995. Mod- elling diseased oyster populations I. Modelling Perkinsiis marimis in- fection in oysters. J. Shellfish Res. 14:121-151. Krantz. G. E. 1982. A Prospectus for the Implementation of Modem Oys- ter Culture Techniques in Maryland. Oyster Hatchery Technology Se- ries. Univ. of Md. Sea Grant Publ. UMSGMAP 82-01. 216 pp. Krantz, G. E. 1989a. A procedural plan to control interjurisdictional trans- fers and introductions of shellfish. Atlantic States Mar. Fisheries Comm. Fish. Mgt. Rep. 13:63. Krantz. G. E. 1989b. Status of MSX and Dermo in Chesapeake Bay oyster stocks in Maryland. Proc. Interstate Seafood Seminar. Md. Dept. of the Environment. Baltimore, pp. 61-78. Krantz. G. E. 1990. Maryland Oyster Population Status Report: 1989 Fall Survey. Maryland Dept. Natural Resources. CBRM-OX-90-1. Annap- olis. Krantz. G. E. 1991 . Maryland Oyster Population Status Report: 1990 Fall Survey. Maryland Dept. Natural Resources. CBRM-0.\-91-l . Annap- olis. Krantz. G. E. 1992. Maryland Oyster Population Status Report: 1991 Fall Survey. Maryland Dept. Nat. Resour. CBRM-OX-92-1. Annapolis. Krantz, G. E. 1993. Maryland Oyster Population Status Report; 1992 Fall Survey. Maryland Dept. Natural Resources CBRM-OX-93-3. Annap- olis. Krantz, G. E. 1995. Maryland Oyster Population Status: 1993 and 1994 Biological Seasons. Maryland Dept. Natural Resources CBRM-OX- 95-1 . Annapolis. Krantz, G. E. & D. Meritt. 1976. An analysis of trends in oyster spat set in the Maryland portion of the Chesapeake Bay. Proc. Natl. Shellfish. Assoc. 67:53-59. Mackin, J. G. 1962. Oyster diseases caused by Denmicxsiuliiim maniuini and other microorganisms in Louisiana. Puhl. Insl. Mar. Sci Univ. Te.xas l:in-229. Maryland Department of Natural Resources. 1993. Maryland Oyster Re- covery Action Plan. Annapolis. Meritt, D. W. 1993. Effects of C liana iriiini. Polydoru websten and Per- kinsus marinus on shell growth, condition index and mortality of the oyster Crassnslrea virginica. Ph.D. Dissertation. Univ. of Maryland, College Park. 167 pp. Otto. S. V. & G. E. Krantz. 1980. More oyster diseases. Proc. Oyster Culture in Maryland. Md. Sea Grant Publ. UM-SG-MAP 81-01:68- 84. Paynter, K. T. & E. M. Burreson. 1991. Effects of Perkinsus mciriniis infection in the eastern oyster, Crassosrrea virginica: II. Disease de- velopment and impact on growth rate at different salinities. J . Shellfish Res. 10(2);425^31. Perkins. F. O. 1988. Structure of protistan parasites found in bivalve mol- luscs. Amer. Fish. Soc. Soc. Spec. Puhl. 18:93-111 Ray. S. M. 1966. A review of the culture method for detecting Dermo- cystiiliiiin marinum with suggested modifications and precautions. Proc. Nail. Shellfi.sh. Assoc. 54:55-69. Roberson. B. S.. T. Li & C. F. Dungan. 1993. Flow cytometric enumer- ation and isolation of immunotluorescent Perkinsus nuinniis cells from estuarine waters. J. Shellfish Res. 12:138. Saunders. G. L., E. N. Powell & D H. Lewis. 1993. A determination of in vivo growth rales for Perkinsus marinus. a parasite of Crassostreu virginica. J. Shellfish Res. 12:229-24(1. Smith, G. F. & S. J. Jordan. 1992. Monitoring Maryland's Oysters: a Comprehensive Characterization of Modified Fall Survey Data 1990- 1991. Md. Dept. Natural Resources CBRM-OX-93-l . Sonial. T. M. 1985. Changes in levels of infection of oysters by Perkinsus niannus. with special reference to the interaction of temperature and salinity upon parasitism. Northwest Gulf Sci. 7:171-174. Turner, H M 1985. Parasites of eastern oysters from subtidal reefs in a Louisiana estuary with a note on their use as indicators of water qual- ity. Estuaries 8:323-325. White, M. E.. E, M. Powell. S M. Ray & E. A. Wilson. 1987. Host to host transfer of Perkinsus marinus in oysters iCrassostrea virainica) populations by the ectoparasitic snail Bimneu impressa (pyramidell- idae). J. Shellfish Res. 6:1-6. Joiinud of Shellfish Research. Vol. 15, No. I. 177-183. 1996. THE ECOLOGICAL ROLE OF THE EASTERN OYSTER, CRASSOSTREA VIRGINICA, WITH REMARKS ON DISEASE VICTOR S. KENNEDY Universin of Maryland Horn Point Environmental Laboratory Box 775. Cambridge. Maryland 21613 .ABSTRACT Historical writings and the presence of pre-Colonial shell middens provide evidence that individual eastern oysters Crassosrrea virginicci (Gmelin. 17911 grew larger and formed more extensive reefs than they do under present-day conditions of harvesting, habitat destruction, and disease. Their diminished abundance has reduced their roles in providing hard substrate, in filtering the estuarine water column, and in affecting energy flow and nutrient flux. The poorly understood interactions among the abundant inhabitants of oyster assemblages have undoubtedly also been affected, although it is not clear how or by how much. The one important difference between the results of mortality from harvesting compared with mortality from disease may be that the shells of disease- stricken oysters remain on the oyster bar to continue to serve as substrate. There may be a connection between stress imposed by overfishing or habitat alteration and susceptibility ot eastern oysters to disease. KEY WORDS: Eastern oyster, ecological role, disease. C. virgimca Every oyster-bed is thus, to a certain degree, a community of living beings, a collection of species, and a massing of individuals, which find here everything necessary for their growth and continuance. ... I propose the word Biocoeno- sis for such a community. Any change in any of the relative factors of a biocoenose produces changes in other factors . . . (Mobius 1877). view I will consider the former abundances of the eastern oyster (also referred to herein as "the oyster" or "oysters") in North America, its ecological role, and some hypotheses as to how that role might have diminished as a result of population depletion due to human activities and disease. INTRODUCTION POPULATION CHANGES OVER TIME Historians of ecology credit Karl Mobius (1877) with being perhaps the first scientist to recognize (at least in print) the exis- tence of assemblages of organisms that interact with one another and with their abiotic environment. He coined the term "bio- coenose" to categorize this "social community" and used as his example the beds formed by the European flat oyster O.strea edit- lis. reporting that (macrofaunal) species diversity was greater on the beds than on the adjacent soft sediments. Presciently he com- mented: In North America the oysters are so fine and cheap that they are eaten daily by all classes. Hence they are now. and have been for a long time, a real means of subsistence for the people . . . but as the number of consumers increases in America the price will also surely advance and then there will anse a desire to fish the banks more severely than hitherto, and if they do not accept in time the unfortunate experience of the oyster culturists of Europe, they will surely find their oyster beds impoverished for having defied the biocoenotic laws (quoted bv Winslow 1881 and Sweet 1941). Unfortunately, Mobius" pessimism was justified. Populations of the eastern oyster Crassostrea viri>inica in North America became overfished (e.g., Ingersoll 1881,Oemler 1894, Haven etal. 1978, Kennedy and Breisch 1983), just as happened to populations of other oyster species elsewhere (see Mobius 1877. Gross and Smyth 1946). In addition, habitat degradation and disease has- tened the decline of some eastern oyster populations. In this re- The ability of unexploited, or relatively unexploited, eastern oysters to flourish over time is demonstrated by two pieces of evidence. The first is the fact that the body size of oysters and the spatial extent of their beds on the New World's East Coast as- tounded the earliest colonial observers. For example. Ingersoll ( 1881 ) cited two accounts of oyster size — one by Wood in 1634 that Charles River. MA. oysters had to be cut in two in order to be eaten, and an undated report by Josselyn that Gulf of Maine oys- ters had to be trisected for the same reason. Similarly. Wharton (1957) quoted the Swiss writer Michel's comment in 1701 that Chesapeake Bay oysters made two mouthfuls. Ingersoll (1881) also repeated Wood's report of 1634 that oysters in the Charles and Mystic Rivers in New England formed reefs that were navigational hazards. Similarly. Michel stated in 1701 that the incredible abun- dances of Chesapeake Bay oysters resulted in reefs so large that ships had to avoid them and that his sloop was grounded on one for 2 hours until the tide returned (Wharton 1957). Finally, early explorers of west Florida reported that oysters formed reefs that broke the water surt'ace (Ingersoll 1881). The second piece of evidence concerning oyster abundances is the presence of coastal shell middens that point to long-term pre- Colonial use of oysters by American Indians. Ingersoll (1 88 1) described such a midden in Damariscotta. ME, that was built by Quoddy Indians. It contained an estimated 226,000 m^ of mostly single shells of eastern oysters, many of them over 30 cm long (supporting the reports quoted above about the size of the edible tissue). Ingersoll (1881) also reported that shells from middens in the St John's River region of Florida were generally larger than 177 178 Kennedy those of living oysters on nearby oyster-producing beds. Similarly, Lunz (1938) found that shells in a South Carolina midden were 61% longer (hmge to bill) and 43% wider than shells of oysters living on nearby commercial beds (the percentages I report aie those corrected by Gunter 1938). Commercial fishing initially depleted the oyster "capital" that so astounded the colonists. For example, in Canada's Maritime Provinces the early fishery grew as improved transportation inland led to expanded markets and increased demand (Stafford 1913). Prices rose, oyster abundances fell, and eventually beds yielded few or no oysters. In some Maritime regions, most of the season's catch was taken in the first day of fishing, with boats massed over beds awaiting the opening hour of the season (Stafford 1913). This description of the rise and fall of Canadian oyster fisheries is generic for the industry on the North American East Coast. By the early 20th Century then, oyster populations had de- clined greatly along the Atlantic Coast of North America. Few oysters remained on the southern shore of Nova Scotia or along the southwest coast of the Gulf of Maine (Stafford 1913. Churchill 1920); the remaining commercial beds in the species' northern range were in warmer Gulf of St. Lawrence waters. Ackerman (1941) reported that there were no recorded landings from Maine or New Hampshire, with the only fished beds in the Gulf of Maine being tiny remnants on the northern shore of Cape Cod. The once plentiful public beds in southern New England had become of minor importance to the fishery, and most oysters were produced by cultivation. Matthiessen (1970) reported that production from North Carolina to east Florida had declined since the turn of the century and that the only region with stable production since about 1900 was that of the Gulf States (west Florida to Texas). In ad- dition to overfishing, destruction of habitat (e.g.. Kennedy and Breisch 1981, Hargis and Haven 1988) and mortalities due to disease (e.g.. Farley 1992) were implicated in these population declines. The diminished abundances of eastern oysters undoubt- edly had a negative effect on their important ecological role in estuaries. THE ECOLOGICAL ROLE OF OYSTERS Oysters as Substrate Oysters play a functional role in providing hard substrate (shell! that is used by members of the oyster biocoenose in the sediment-dominated environment of an estuary. Dauer et al. (1982) demonstrated in lower Chesapeake Bay that clumps of scrubbed oyster shells on experimental plots became associated with significantly higher densities of invertebrate macrofauna. both on the shells themselves and in the soft sediments between clumps, than occurred in untreated control plots. The shells pro- vided attachment space for tubiculous suspension feeders (primar- ily polychaetcs) whose tubes in turn became attachment space for additional epifauna. Shells of barnacles and byssate mussels that attach to oyster shell also serve as sources of hard substrate for epifaunal foulers (personal observations). The role of substrate provider is enhanced by the irregularity of oyster shell surfaces. The shell veneer of an oyster bed displays a diversity of configurations; that is. there can be single shells or shell pieces, articulated shell pairs, and clumps or clusters of at- tached shells scattered on the bed. producing a complex, three- dimensional structure. Bahr (1974) estimated that 1 m" of inter- tidal oyster reef in Georgia provided 50 -t- nr of available surface for epifauna to exploit. Community Structure and Dynamics The ecological role of oyster communities in the economy of estuaries is influenced by the structure (physical, biological) and function of oyster beds and their components. As to physical struc- ture. Kennedy and Sanford (unpublished observation) have exam- ined early descriptions of beds of eastern oysters in North Amer- ica, their morphology, and the influence of environmental factors, especially water circulation, on physical structure. Here 1 examine biological structure briefly and then consider the topic of function. Note that in general on the East Coast of North America, oyster beds from Chesapeake Bay north are subtidal, with beds from South Carolina south being mainly intertidal and North Carolina representing a transition zone of intertidal and subtidal habitat (Kennedy and Sanford, unpublished observation). In the Gulf of Mexico, most oyster beds are subtidal. The spatial distribution of oysters on a bed is not well under- stood, nor are the factors influencing such distributions. In perhaps the only study of its kind, Powell et al. (1987) examined the small-scale spatial distribution of eastern oysters on 1 1 beds on the Texas coast. The size categories used were: small, <2 cm; me- dium, 2.1-5.0 cm; large, >5 cm. Small oysters were patchily distributed, presumably as a result of contagious settlement or. perhaps, differential mortality. Patch size decreased with increas- ing oyster size as oysters became less contagiously distributed. Clumps with relatively few large oysters tended to be found im- mediately adjacent (<12 cm apart) to clumps with relatively many large oysters. Powell et al. ( 1987) attributed this latter finding to possible superior competition for food by the more abundant and larger clumped oysters, with such competition also enhancing the negative effects of predators and disease on the adjacent clump of smaller and fewer oysters. Another poorly understood aspect of community dynamics is that of interactions among the numerous associates on oyster beds. To begin with, only a few studies (Table 1) have examined the faunal composition of oyster assemblages in eastern North Amer- ica. The most southerly investigations identified from 31 to 155 invertebrate taxa (not all identified to species) associated with oyster beds. Studies in Chesapeake Bay (Larsen 1985) and Dela- ware Bay (Maurer and Watling 1973) yielded from 129 to 138 invertebrate taxa; Frcy's ( 1946) cursory collection of invertebrates produced about 41 species in the Potomac River, a tributary of Chesapeake Bay. Wells' (1961) extensive examination of oyster beds in North Carolina yielded 284 taxa. with 55-65 taxa collected from predominantly intertidal beds (he did not provide details on intertidal versus subtidal diversities). The macrofaunal inverte- brates in all studies represented a mixture of filter and deposit feeders, as well as mobile predators and a variety of tunicates and benthic fish. The four studies that measured faunal density pro- duced estimates of thousands of individuals per square meter (Ta- ble 1), with Larsen ( 1985) recording a high of 125.573 individuals m"" in the James River. VA. The cited studies are only comparable qualitatively because of differences in sampling methods (sampling gear, screen sizes) and sampling intensity (spatially, temporally) and because community structure can change with season and along the salinity gradient (Dame 1979, Dauer et al. 1982, Larsen 1985). Nevertheless, the impression is one of relatively high species diversity and (espe- cially) high faunal abundances among invertebrate macrofauna on oyster beds. For example. Larsen (1985) found that species rich- ness on James River oyster beds was generally similar to values The Ecological Role of the Eastern Oyster 179 TABLE 1. Number of invertebrate taxa (not all identifled to species), species' biomass, and faunal density (individuals m ^1 associated with beds of C. virginica in USA (S = subtidal; I = intertidal; - = not measured. Location abbreviations refer to states in USA). Geographic Location Number Biomass Faunal Location of Beds of Taxa (g m ^) Density Reference Northern Gulf of Mexico S/I 155 — — Kilgen & Dugas 1989 Redfish Bay. TX S/I — 479 — Copeland & Hoese 1966 Apalachicola Bay. FL s 90 — — Pearse & Wharton 1938" Crystal River. FL I 31 253 6.200 Lehman 1974 Georgia I 42 705 24,747 Bahr 1974 North Inlel. SC I 37 214 2.476- 4.077 Dame 1979 Newport River. NC S/I 284 — — Wells 1961 James River. VA s 138 — 5.757- 57.857" Larsen 1985 Potomac River. MD s 41'- — — Frey 1946 Delaware Bay. DE s 129 — — Maurer & Watling 1973 " My estimate of benthic species, excluding proto/oa and parasites. "' Mean faunal density. ' My estimate, excluding internal parasites and pelagic species. reported for mesohaline soft-bottom assemblages but that mean densities of macrofauna were much higher on the beds. There is scope for much more research on the influence of oyster aggregations on biodiversity. Increased faunal biodiversity has been associated with clumps of mytilid species on both soft- sediment shores (e.g., Dittmann 1990, Thiel and Demedde 1994) and hard-sediment shores leg.. Lintas and Seed 1994 and refer- ences therein). Breitburg (unpublished observation) reported that recruitment rates of (he naked goby. Gobiosoma base, are 10-100 times higher than recruitment rates recorded for coral reef fish or other temperate reef fish species. The modification of water flow by oyster shell on the beds enhances aggregation of naked goby larvae (Breitburg ct al.. 1995). Juvenile striped bass [Morone saxatilis) are abundant over oyster beds in Chesapeake Bay. per- haps preying on naked goby larvae (Breitburg. unpublished ob- servation). Interactions among the various species on oyster beds are not well understood because research has been limited. In one in- stance, Ortega (1981) examined intertidal oyster assemblages in North Carolina. At a wave-exposed coastal site, C . virginica was uncommon and the scorched mussel Brachiodontes ( = Brachi- donles) e.xii.stus was dominant in the lower and middle intertidal. Ortega (1981) attributed this pattern to low colonization rates by the oyster, its sensitivity to wave shock, and overgrowth by the mussel. At a protected site, she found that the oyster dominated the low and middle intertidal and attributed this pattern to higher colonization rates than those of barnacles, greater tolerance of heat and desiccation (thought to be deleterious to the mussel), and rapid growth. She proposed that the absence of predatory gastropods in the protected intertidal region also helped the oyster dominate the habitat. The partitioning of oyster bed habitat by various mud crabs (family Xanthidae) has received limited attention. Their distribu- tion on oyster beds was examined in Delaware Bay by McDermott and Flower (1952) and in Chesapeake Bay by Ryan (1956), and Day and Lawton (1988) demonstrated that three species of mud crabs preferred broken oyster shell over four other types of sub- strate. However, we barely understand how these small crusta- ceans make use of the three-dimensional shell structures on oyster beds. In South Carolina, Panopeus herbstii (an oyster predator) and Eurypanopeus depressus co-occur on intertidal beds, with the smaller E. depressus generally restricted to the narrower spaces among shells (McDonald 1982). Meyer (1994) confirmed this finding for these same two species on intertidal beds in North Carolina. He also reported that P . herbstii (especially, large indi- viduals) is more common under shell on and within the oyster bed compared with E. depressus, which is common in oyster-shell clusters that pro|ect above the surface of the bed. Winter temper- atures are associated with a shift in distribution by E. depressus and small P . herbstii from shell clusters to subsurface shell, pre- sumably away from the more extreme air temperatures (Meyer 1994). In Florida, adult and juvenile specimens of another species of mud crab, Panopeus obesus, are concentrated in the high in- tertidal zone on oyster reefs, whereas adult Panopeus sinipsoiu occupy the lower intertidal. with juvenile P. simpsoni found ho- mogenously over the reef (Mcnendez 1987). Finally, in Georgia, ribbed mussels. Geukensia demissa. that attach to the exterior of oyster clumps experience nearly four times the mortality from attack by P . herbstii than do mussels within the clumps" interstices (Lee and Kneib 1994). The partitioning of habitat by mud crabs on subtidal reefs is unknown, save for the report that in Delaware Bay the abundant Dyspanopeus sayi ( = Neopanope texana sayi) is common in the bodies of red-beard sponge {Microciona prolifera) that attaches to oyster shell (McDermott and Flower 1952). In terms of other oyster-associated invertebrates, habitat partitioning probably oc- curs among epifauna like hydroids. bryozoans. and barnacles. Such ecologically interesting problems can be investigated readily on oyster beds. At a scale smaller than that of the oyster bed. the landscape microecology of eastern oyster shell is an area awaiting detailed study. Korringa (1951) described the assemblage of plants and invertebrate animals that used the shell of O. edulis as a habitat. No such detailed study has been performed on the shell of eastern oysters. However, Osman et al. (1989) examined the interplay among fouling organisms and settling or settled eastern oysters. They found that growth of newly settled spat was inhibited and survival was reduced in association with an assortment of sessile 180 Kennedy invertebrates that co-occurred on the cultch, with competition for food apparently being an important limiting factor (Zajac et al. 1989). Additional experiments (Osman et al. 1990) showed that ontogenetic changes in trophic relationships led to complex inter- actions among spat and associated predators and competitors. In addition to the influence of such biological factors, physical fac- tors such as depth in the water column may influence community structure, as was found by Hirata (1987) for a Japanese species of oyster. Crassostrea iiippoiw. and some of its associates. Oyster beds not only provide a refuge from extreme environ- mental conditions as noted for mud crabs above, but also a refuge from predation. and even as a waiting place for predators. For example, Posey et al. ( 1995) reported that oyster beds serve as a refuge for grass shrimp. Paleomoiietes pugio. and other decapods from predation by predatory fish. In addition, some species of predators remain in moist habitat on the reef during low water, moving off the reef in high water to forage on the adiacent soft-sediment benthos. The depletion of oyster beds has potentially major effects on species in addition to those using the shell as a substrate or a refuge from extreme temperatures or predation. Pea crabs. Pinnotheres ostreum. live within the shell cavity of oysters (e.g., Christensen and McDermott 1958). There appear to be no long-term data on abundances of pea crabs in relation to oyster abundances, so 1 cannot speculate on the effects of depleted oyster abundances. It is possible that pea crabs have sufficient alternative bivalve hosts (e.g., Sandifer and Van Engel 1970) to allow their populations to persist even as oysters have become less common. A smiilar lack of data hinders attempts to correlate abundances of two fish that feed on oysters (among other prey), the oyster toadfish (Opsanus tau) and the sheepshead (Archosargus probatoccphahis). with changes in oyster abundances. At the microscale level, oyster gametes (and, presumably, pe- lagic gametes of associated organisms) serve as food for micro- heterotrophs and metazoan suspension feeders, with oyster sperm rapidly ingested by microprotozoans (Galvao et al. 1989). These investigators estimated that over 50% of oyster spemi released in a salt marsh could be ingested by the resident population of mi- crobial grazers. In the absence of understanding of energy budgets of microbial food webs, it is difficult to say how significant is the loss of gametes from residents of oyster beds when populations of those residents have been depleted by human activities or disease. Energy Flow and SutrienI Flux The difficult task of modelling the energetics of oyster beds has rarely been undertaken, with the bulk of such research on C. virginica occurring in South Carolina (Dame 1972, 1976. Dame and Patten 1981) and Georgia (Bahr 1976). Dame (1976) found that intertidal oysters in South Carolina had the greatest population production (P), assimilation (A; energy flow), and net growth efficiency (P/A * 100) of nine intertidal molluscs studied to that date. Dame and Patten (1981) reported that an intertidal reef in South Carolina consumed about 15,000 Kcal m"~ y"'. Bahr (1976) and Bahr and Lanier (1981) estimated that the total com- munity respiration of an intertidal reef in Georgia was 27,000 Kcal m^- y~'. These values are very high for natural heterotrophic systems (Dame and Patten 1981). Similar studies need to be per- formed for subtidal beds in more northerly climes and in the Gulf of Mexico to provide a comparison of energy flow under constant submergence. In terms of nutrient flux, oyster beds are thought to play sig- nificant roles in habitat in which they abound. For example, on intertidal beds in South Carolina. Dame et al. (1984) measured release rates of ammonia that were comparatively higher than rates measured in other coastal and estuarine habitats and proposed that C. virginica was important in material cycling. Jordan (1987) reported that production of feces and pseudofeces by subtidal C. virginica in Chesapeake Bay played an important role in sedimen- tation and remineralization. It may be instructive that an intertidal bed of the Pacific oyster Crassostrea gigas in France yielded ev- idence of complex chemical interactions (oxygen consumption, fluxes of nutrients) depending upon the available biomass of oys- ters (Boucher-Rodoni and Boucher 1990). Much research is un- derway into the role of beds of other bivalves in nutrient exchanges (e.g., Asmus et al. 1995) and there is scope for similar research into the role played by intertidal and subtidal oyster beds. Oysters and Food Webs Limited attention was paid to the ecological role of oysters and associated macrofauna in food webs until Newell (1988) empha- sized the extent to which suspension-feeding oysters link pelagic and benthic food webs. He proposed that Chesapeake Bay"s ex- tensive oyster populations before 1870 had the potential in the summer to filter the Bay's water column in less than a week, whereas the curtently depleted populations may take over 46 weeks. Unless some other suspension-feeding group(s) made up the difference (and there is no evidence that they did), this decline in filtering capacity would lessen the grazing pressure on phyto- plankton populations. Newell (1988) proposed that this change would shift Bay food webs from a benthic-dominated to a pelagic- dominated mode, with the sea nettle Chiysaora cjuuuiuecirrha assuming a major role in controlling energy flow. There is evidence to support Newell's ( 1988) claims. For ex- ample. Ulanowicz and Tuttle (1992) used a simple network- analysis model of mesohaline Chesapeake Bay to ask ""what if oyster populations became more abundant?" By decreasing oyster harvest per unit biomass by 23% in their model, they were able to predict an increase of 150% in oyster biomass and an 89% de- crease in gelatinous zooplankton (which includes sea nettles). In addition, the model predicted a 29% increase in benthic diatoms. This prediction is important because Cooper and Brush (1993) have since demonstrated that the ratio of centric diatoms (usually planktonic and associated with eutrophic environments) to pennate diatoms (usually benthic and from clear waters) has increased many-fold over the past half-century in the Bay. It is always difficult to perform hindcasting experiments on complex ecosystems. It can be argued that the model of Ulanowicz and Tuttle (1992) is simplistic and subject to making unrealistic predictions. Further, a decline in benthic diatoms can result from light limitation caused by increasing turbidity from sediment run- off in the wake of human land-clearing activities since Colonial times. The lack of available data on population changes over time in the various benthic and pelagic components of the Bay's eco- systems can make the exercises of Newell (1988) and Ulanowicz and Tuttle (1992) seem too speculative. Serendipitously, some natural experiments involving invasions of aquatic ecosystems by exotic bivalves have demonstrated that bivalves can indeed have major effects when their populations expand. For example, Cohen et al. (1984) found that high densi- ties of the introduced Asiatic clam Corbiciita fluminea in the fresh- The Ecological Role of the Eastern Oyster water Potomac River. MD, coinelded with a region in which phy- toplantcton abundances were 40-60'^ below abundances immedi- ately upstream or downstream of the clams. Similarly. Alpine and Cloern (1942) documented declines in phytoplankton abundance as populations of the exotic Asian clam PoUimocorbula amitreiuis increased in the estuary of San Francisco Bay. The invasion of North America's Great Lakes by the zebra mussel Dreissena poty- morplni has led to many reports of depicted phytoplankton popu- lations and increased water transparency in the lakes (e.g.. Hol- land 1993). Further. Stcuart and Haynes (1994) provide evidence that the zebra mussels have modified energy pathways in Lake Ontario by deposition of feces and pseudofeces. The mussel col- onies also appear to enhance substrate complexity, and abundance and diversity of benthic macroinvertebrates in the vicinity of the colonies are higher than in areas without colonies. As noted above, it is not clear that other suspension-feeding organisms have replaced oysters where the latter once thrived (Newell 19SS). Mobius (IS77) reported that populations of flat oysters that declined in western European waters were often re- placed b\ blue mussels. M\iilii.\ alulis. and cockles. Caidiiim sp The Atlantic rangia clam H(iiii;ui ciint'uia has increased in abun- dance in soft sediments in oligohaline and upper mesohaline Ches- apeake Bay (Hopkins and Andrews 1969). but there are no data on how its abundances and filtering capacities resemble those of the eastern oyster. In addition, its populations fluctuate in the upper Bay o\er time (pers. obs.). Interactions among aquatic herbivores and primary producers are complex and of widespread interest (e.g., Prins et al. 1995). The role played by bivalves other than the oyster, as well as other suspension feeders, in the economy of the Bay and elsewhere is a topic that requires detailed examination. THE ECOL()(n OK DISEASE IN OUSTERS In addition to overfishing and habitat degradation, disease has also taken its toll on populations of eastern oysters (Ford and Tripp 1996). beginning in recorded memory in Malpeque Bay. Prince Edward Island. Needier ( 1931 ) slated that oyster stocks in Malpe- que Bay had been greatly overfished by the turn of the 20th Cen- tury, with landings in 1914 being 10C{ of those in 1882. the historical peak harvest. Subsequently, the onset of Malpeque Bay Disease (thought to have been introduced with ovsters imported to bolster the local fishery) caused high mortalities in 1913 and 1916. leading to no measurable landings in 1918. In a similar manner, overfishing of Delaware Bay and Chesapeake Bay stocks led to diminished returns in the first half of the 2()th Century, with sub- sequent depletion being made worse by MSX and dermo diseases (Andrews 1968. Haven et al. 1978. Haskin and Ford 1982). Disease can be debilitating to oysters by inhibiting growth, lowering condition, and disrupting filtering activities, depending upon the particular etiological agent (see Ford and Tripp 1996 and Paynter 1996 for reviews). Negative but nonlethal effects on phys- iological activities can indirectly influence the oysters' roles in energy flow and nutrient flux. However, our knowledge of these matters is not sufficient to allow for predictions to be made about the extent of such nonlethal effects. In most ways, depletion of oyster populations by the lethal effects of disease is similar in effect to depletion by human activ- ities. That is. the extent of the biological activity of oysters (shell and gamete production, particle filtration and deposition, nutrient flux) on the bed is reduced as oyster abundances decline. Again, it IS not yet possible to quantify such changes with any accuracy. Beyond the physiological effects noted above, perhaps the major difference in possible ecological effects between disease and an- thropogenic factors is that the shells of oysters killed by disease remain on the bed ( m contrast to the removal of shell from beds by harvesting). This allows them to continue to function as substrate or as a refuge for small invertebrates and fish as well as their eggs and larvae. Finally, although Farley (1992) associated most mass mortal- ities in oysters w ith transfers of infected animals, the occurrence of disease in the wake of overfishing as noted above is provocative. Is it possible that susceptibility to disease in oysters can be en- hanced in some instances by heavy fishing activity? Is there a deleterious level of stress in oysters on physically disturbed beds that have been scraped to just above the estuarinc bottom? Gross and Smyth (1946) evaluated the history of declining oyster populations, especially with regard to O. ediilis. and pro- posed that overfishing leads to loss of genetic variability and re- duced adaptability to long-term environmental changes. Laird ( 1961 ) hypothesized that when physical factors (e.g.. temperature, salinity, tidal factors) led to a physiologically unfavorable envi- ronment for eastern oysters, they became susceptible to disease (he was considering hcxamitosis caused by Hcxainita injlatu). He pro- posed that the proximity of the sediment-water interface led to stress and depressed resistance in eastern oysters. Lee et al. ( 1995) suggest that a fungus associated with a species of yew tree may have been responsible for the precipitous decline in abundance of the yew. They hypothesize that the fungus was a relatively benign associate of the yew as an endophyte until environmental changes caused by unregulated forestry practices stressed the yew and ren- dered it vulnerable to attack by the fungus. There is evidence that stressors can influence prevalence and intensity off. lunnniis disease in eastern oysters. Laboratory stud- ies have shown that oysters are increasingly infected by P nuiii- nii\ in the presence of pollutants of various kinds (Wilson et al. 1990. Chu and Hale 1994. see also Paynter 1996). Whether this is due to increased susceptibility on the part of the host oyster (e.g.. from physiological stress or suppressed immune systems) or to enhanced virulence of the parasite is not clear. In contrast to these observations and speculations, dermo disease has been observed in oysters held in trays in the upper water column (R. Newell. Horn Point Environmental Laboratory, pers. conim.). Presumably these oysters were less subject to hypoxia and sedimentation, with any attendant deleterious effects, yet they became infected. Clearly, the subject of environmental influences on suscepti- bility to disease and any role that stress from harvesting and habitat alteration might play in susceptibility require additional research. I propose that, because disease is a natural component of biolog- ical systems, it is unlikely to depress populations of the eastern oyster for long (time measured in decades or centuries), except in concert with the stressors imposed by humans. Meanwhile, until oysters develop a natural resistance to disease or perhaps until advances in biotechnology produce such resistance, managers of eastern oyster fisheries will need to implement recommendations such as those of Andrews and Ray (1988) in order to manage around the effects of P. miiriiuis. ACKNOWLEDGMENTS 1 thank G. Abbe. J. Kracuter. R. Osman. and an anonymous reviewer for their comments on earlier drafts of this paper. 182 Kennedy LITERATURE CITED Ackerman, E. A. 1941. New England's Fishing Indiisln. University of Chicago Press, Chicago. Ilinois. 303 pp. Alpine. A. E. & J. E. Cloem. 1992. Trophic interactions and direct phys- ical effects control phytoplankton biomass and production in an estu- ary. Limnol. Oceanogr. 37:946-955. Andrews. J. D. 1968. Oyster mortality studies in Virginia. VII. Review of epizootiology and origin oi Minchinia nelsoni. Proc. Nail. Shellfish. Assoc. 58:23-36. Andrews, J. D. & S. M. Ray. 1988. Management strategies to control the disease caused by Perkinsus marinus. Amer. Fish. Soc. Spec. Piihl. 18:257-264. Asmus. H.. R. M. Asmus & G. F. ZubiUaga. 1995. Do mussel beds intensify the phosphorus exchange between sediment and tidal waters? Ophelia 4\:}1~55. Bahr. L. M. 1974, Aspects of the structure and function of the intertidal oyster reef community in Georgia. Doctoral dissertation. Universitv of Georgia. Athens. Georgia. 149 pp. [Not seen: quoted In Bahr & Lanier 1981]. Bahr. L. M., Jr. 1976. Energetic aspects of the intertidal oyster reef com- munity at Sapelo Island. Georgia (USA). Ecology 57:121-131. Bahr. L. M. & W. P. Lanier. 1981. The ecology of intertidal oyster reefs of the South Atlantic coast: a community profile. U.S. Fish and Wild- life FWS/OBS/81.15. Washington, DC. 105 pp. Boucher-Rodoni. R. & G. Boucher, 1990 In siru study of the effect of oyster biomass on bcnthic metabolic exchange rates. Hvtlrobiologia 206:115-123. Breitburg. D. L.. M. A. Palmer & T. Loher. 1995. Larval distnbutions and the spatial patterns of settlement of an oyster reef fish: responses to flow and structure. Mar. Ecol. Prog. Ser. 125:45-60. Christensen. A. M. & J. J. McDermott. 1958. Life-history and biology of the oyster crab. Pinnotheres oslreiim Say. Biol. Bull. 1 14:146-179. Chu. F.-L. E. & R. C. Hale. 1994. Relationship between pollution and susceptibility to infectious disease in the eastern oyster. Crassosirea virginica. Mar. Environ. Res. 38:243-256. Churchill, E. P., Jr. 1920. The oyster and the oyster industry of the At- lantic and Gulf coasts. Bureau of Fisheries Document 890. .Appendix VIII, Report U.S. Commissioner of Fisheries for 1919:1-51 Cohen, R. R. H., P. V. Dresler, E J. P. Phillips & R. L. Cory. 1984 The effect of the Asiatic clam, Corbicula fluminea. on phytoplankton of the Potomac River, Maryland. Limnol. Oceanogr. 29:170-180. Cooper. S. R. & G. S. Brush. 1993. A 2,500-year history of anoxia and eutrophication in Chesapeake Bay. Estuaries 16:617-626. Copeland, B. J. & H. D. Hoese. 1966. Growth and mortality of the Amer- ican oyster, Crassosirea virginica. in high-sahnity shallow bays in central Texas. Publ. Inst. Mar. Sci. Univ. Te.ws 1 1:149-158. Dame. R. F. 1972. The ecological energies of growth, respiration and assimilation in the intertidal American oyster Crassostrea virginica. Mar. Biol. 17:243-250. Dame. R. F. 1976. Energy flow in an intertidal oy.ster population. £.««- arine Coastal Mar. Sci. 4:243-253. Dame, R. F. 1979. The abundance, diversity and biomass of macro- benthos on North Inlet, South Carolina, intertidal oyster reefs. Proc. Natl. Shellfish. Assoc. 69:6-10. Dame, R. F. & B. C. Patten. 1981. Analysis of energy flows in an inter- tidal oyster reef. Mar. Ecol. Prog. Ser. 5:115-124. Dame. R, F.. R. G. Zingmark & E. Haskin. 1984. Oyster reefs as pro- cessors of estuarine materials. J. Exp. Mar. Biol. Ecol. 83:239-247. Dauer, D. M., G. H. Tourtellotte & R. M. Ewing. 1982. Oyster shells and artificial worm tubes: the role of refuges in structuring benthic communities of the lower Chesapeake Bay. Int. Rev Gesamtcn H\- drobiol. 6 '661-677. Day, E. A, & P. Lawton. 1988. Mud crab (Crustacea: Brachyura: .Xan- thidae) subs ate preference and activity. J. Shellfish Res. 7:421^26. Ditlmann, S. 1990. Mussel beds — amensalism or amelioration for inter- tidal fauna? Helgol. Meeresunters. 44:335-352. Farley, C. A. 1992. Mass mortalities and infectious lethal diseases in bivalve molluscs and associations with geographic transfers of popu- lations. In: A. Rosenfield & R. Mann (eds.). Dispersal of Living Organisms into Aquatic Ecosystems. Maryland Sea Grant, College Park, Maryland, pp. 139-154. Ford. S. E. & M. R. Tripp. 1996. Diseases and defense mechanisms. In: A. F. Eble, V. S. Kennedy & R. I. E. Newell (eds.). The Eastern Ovster, Crassostrea virginica. Maryland Sea Grant, College Park, Maryland, pp, 581-660. Frey, D, G. 1946. Oyster bars of the Potomac River. U.S. Fish Wildl. Ser\: Spec. Sci. Rep. 32:1-93. Galvao, H. M., A. T. Fritz & R. Schmaljohann. 1989. Ingestion of ga- metes by protists: fate of surplus reproductive energy in the sea. Mar. Ecol. Prog. Ser. 51:215-220. Gross, F. & J. C. Smyth. 1946. The decline of oyster populations. Nature 157:.540-542. Gunter. G. 1938. Comments of the shape, growth, and qualitv of the American oyster. Science 88:546-547. Hargis, W. J., Jr. & D. S. Haven. 1988. Rehabilitation of the troubled oyster industry of the lower Chesapeake Bav, J. Shellfish Res. 7:271- 279, Haskin. H. H. & S. E. Ford. 1982. Haplosporidium nelsoni (MSX) on Delaware Bay seed oyster beds: a host-parasite relationship along a salinity gradient. J. Inverlebr. Pathol. 40:388-405. Haven. D. S.. W. J. Hargis, Jr. & P. C. Kendall. 1978 The Ovster In- dustry of Virginia: Its Status. Problems and Promise. Virginia Institute of Marine Science Special Report 168, Gloucester Point, Virginia. Hirata, T, 1987. Succession of sessile organisms on experimental plates immersed in Nabeta Bay, Izu Peninsula, Japan, II Succession of in- vertebrates. Mar. Ecol. Prog. Ser. 38:25-35. Holland, R, E, 1993. Changes in planktonic diatoms and water transpar- ency in Hatchery Bay, Bass Island area, western Lake Erie since the establishment of the zebra mussel. J. Great Lakes Res. 19:617-624. Hopkins, S. H. & J. D. Andrews. 1969. Rangia cuneata on the east coast: thousand mile range extension, or resurgence? Science 167:868-869. Ingersoll. E. 1881. The oyster industry. In: G. B. Goode (ed.l. The His- tory and Present Condition of the Fishery Industries. Department of the Intenor. Tenth Census of the United States. Washington. DC. pp. 1-252. Jordan. S, J. 1987. Sedimentation and remineralization associated with biodeposition by the American oyster Crassostrea virginica (Gmelin). Doctoral dissertation. University of Maryland, College Park, Mary- land. 203 pp. Kennedy. V. S. & L. L. Breisch. 1981. Maryland's Ovsters: Research and Management. Maryland Sea Grant Publication UM-SG-TS-8I-04 College Park, Maryland. 286 pp. Kennedy, V. S. & L. L. Breisch. 1983. Sixteen decades of political man- agement of the oyster fishery in Maryland's Chesapeake Bay. J . En- viron. Manage. 16:153-171. Kilgen. R. H. & R. J. Dugas. 1989. The ecology of oyster reefs of the northern Gulf of Mexico: an open file report. U.S. Fish & Wildlife Svc, Natl. Wetlands Resch. Ctr. Open File Report 89-03. Washing- ton, DC. Komnga, P. 1951. The shell of Ostrea edulis as a habitat. Arch. Neerl. Zool. 10:32-152. Laird, M. 1961. Microecological factors in oyster epizootics. Can. J. Zool. 39:449-485. Larsen, P. F. 1985. The benthic macrofauna associated with the oyster reefs of the James River estuary, Virginia, U.S.A. Int. Rev. Gesamtcn Hydrobiol. 70:797-814. Lee, S. Y. & R. T. Kneib. 1994. Effects of biogenic structure on prey consumption by the xanthid crabs Eurytium limosum and Panopeus herbslii in a salt marsh. Mar. Ecol. Prog. Ser. 104:39-47. Lee, J. C, X. Yanc, M. Schwanz. G. Strobel & J. Clardv. 1995. The The Ecological Role of the Eastern Oyster 183 relationship between an endangered North American tree and an en- dophytic fungus. Chem. & Biol. 2:721-727. Lehman, M. E. 1974. Oyster reefs at Crystal River. Florida and their adaptation to thermal plumes. Doctoral dissertation. University of Flor- ida, Gainesville. Flonda. 197 pp. [Not seen; quoted in Bahr & Lanier 1981]. Lintas, C. & R. Seed. 1994. Spatial variation in the fauna associated with Mytitus edulis on a wave-exposed rocky shore. J . Mollusccm Slud. 60:165-174, Lunz, G. R., Jr. 1938. Comparison between pre-Colonial and present-day oysters. Science 87:367. Matthiessen, G. C. 1970. A Review of Oyster Culture and the Oyster Industry in North America. Contribution No. 2528 of the Woods Hole Oceanographic Institution. Woods Hole. Massachusetts Maurer. D. & L. Watling. 1973. Studies on the oyster community in Delaware: the effects of the cstuarine environment on the associated fauna. Int. Rev. Gesamten Hydrolnol. 58:161-201. McDermott. J. J. & F. B. Flower. 1952. Preliminary studies of the com- mon mud crab on oyster beds of Delaware Bay. Proc. Natl. Shellfisli. Assoc. 1952:47-50. McDonald. J. 1982. Divergent life history patterns in the co-occurring intertidal crabs Panopeus hert>stii and Eurypanopeus depressus (Crus- tacea: Brachyura: Xanthidae). Mar. Ecol. Prog. Ser. 8:17.3-180. Menendez. R. J 1987 Vertical zonation of the xanthid mud crabs Pan- opeus obesus and Panopeus simpsoni on oyster reefs. Bull. Mar. Sci. 40:73-77. Meyer, D. L. 1994. Habitat panitioning between the xanthid crabs Pan- opeus herbstii and Eurypanopeus depressus on intertidal oyster reefs (Crassostrea virginica) in southeastern North Carolina. Estuaries 17: 674-679. Mobius, K. 1877. An oyster-bank is a biocoenose, or a social community. Original in Die Ausrer und die Austerwirtlischaft . Verlag von Wie- gundt. Hemper and Parey. Berlin. English translation by H. J. Rice ( 1883) in Report U.S. Commission of Fish and Fisheries for 1880. Part 8. Appendix H: 683-751. Excerpts reprinted in E. J. Komiondy (ed.). 1965. Readings in Ecology. Prentice-Hall. Englewood Cliffs. New Jersey, pp. 121-124. Needier A. W. H. 1931 . The oysters of Malpeque Bay. Biol. Board Can- ada Bull. 21:1-36. Newell. R. I. E. 1988. Ecological changes in Chesapeake Bay: Are they the result of overharvesting the American oyster, Crassostrea virgin- ica? In: M. P. Lynch & E. C. Krome (eds.). Understanding lite Es- tuary: Advances in Chesapeake Bay Research. Chesapeake Research Consortium Publication 129, Solomons, Maryland, pp. 536-546. Oemler. A. 1894. The past, present, and future of the oyster industry of Georgia. Bull. U.S. Fish Comm. 13:263-272. Ortega, S. 1981. Environmental stress, competition and dominance of Crassostrea virginica near Beaufort. North Carolina. USA. Mar. Biol. 62:47-56. Osman, R. W., R. B. Whitlatch, R J. Malatesta & R. N. Zajac. 1990. Ontogenetic changes in trophic relationships and their effects on re- cruitment. In: M. Barnes & R. N. Gibson (eds.). Trophic Relation- ships in the Marine Environment . Aberdeen University Press, Scot- land, pp. 117-129. Osman, R. W., R. B. Whitlatch & R. N. Zajac. 1989. Effects of resident species on recruitment into a community: larval settlement versus post- settlement mortality in the oyster Crassostrea virginica. Mar. Ecol. Progr. Ser. 54:61-73. Paynter, K. 1996. The effects oi Perkinsus marimis infection on physio- logical processes in the eastern oyster, Crassostrea virginica. J. Shell- fish Res. 15:119-125. Pearse, A. S. & G. W. Wharton. 1938. The oyster "leech," Stylochus inimicus Palombi, associated with oysters on the coasts of Florida. Ecol. Monogr. 8:605-655 Posey, M, H., C. M. Powell & T. D. Alphin. 1995. Use of oyster reef habitat by fish and decapods. Abstract. Estuarine Research Federation Biennial Meeting. Corpus Christi. Texas. Powell. E. N., M. E. White. E. A. Wilson & S. M. Ray. 1987. Small- scale spatial distribution of oysters (Crassostrea virginica) on oyster reefs. Bull. Mar. Sci. 41:835-855. Pnns, T. C. V. Escaravage, A. C. Smaal & T. C. H. Peeters. 1995. Nutrient cycling and phytoplankton dynamics in relation to mussel grazing in a mesocosm experiment. Ophelia 41:289-315. Ryan, E. P. 1956. Observations on the life histories and the distribution of the Xanthidae (mud crabs) of Chesapeake Bay. Am. Midi. Nat. 56: 138-162. Sandifer. P. A. & W. Van Engel. 1970. Modiolus demissus: a new host for the oyster crab Pinnotheres oslreum in Virginia. Veliger 13:145- 146 Stafford. J. 1913. The Canadian Oyster. Its Development. Environment and Culture. The Mortimer Co.. Ltd. Ottawa. Canada. 147 pp. Stewart. T. W. & J. M. Haynes. 1994. Benthic macroinvertebrate com- munities of southwestern Lake Ontario following invasion of Dreis- sena. J. Great Lakes Res. 20:479-493. Sweet. G. 1941. Oyster conservation in Connecticut: past and present. Geogr. Rev. 31:591-608. Thiel. M. & T. Demedde. 1994. Recruitment of shore crabs Carcinus maenas on tidal flats: mussel clumps as an important refuge for juve- niles. Helgol. Meeresunlers. 48:321-332. Ulanowicz, R. E. & J. H. Tuttle. 1992. The trophic consequences of oyster stock rehabilitation in Chesapeake Bay. Estuaries 15:298-306. Wells. H. W. 1961. The fauna of oyster beds, with special reference to the salinity factor. Ecol. Monogr. 31:239-266. Wharton. J. 1957. The Bounty of the Chesapeake. Fishing in Colonial Virginia. University Press of Virginia. Charlottesville. Virginia. 78 pp. Wilson. E. A.. E. N. Powell. M. A. Craig. T. L. Wade & J. M. Brooks. 1990. The distnbution of Perkinsus marinus in Gulf coast oysters: Its relationship with temperature, reproduction, and pollutant body bur- dens. Int. Rev. Gesamten Hydrobiol. 75:533-550. Winslow. F. 1881. Deterioration of American oyster-beds. Popular Sci- ence Monthly 20:29-43, 145-156. Zajac, R. N., R. B. Whitlatch & R. W. Osman. 1989. Effects of inter- specific density and food supply on survivorship and growth of newly settled benthos. Mar. Ecol. Prog. Ser. 56:127-132. THE NATIONAL SHELLFISHERIES ASSOCIATION The National Shellfisheries Association (NSA) is an international organization of scientists, manage- ment officials and members of industry that is deeply concerned and dedicated to the formulation of ideas and promotion of knowledge pertinent to the biology, ecology, production, economics and man- agement of shellfish resources. The Association has a membership of more than 1000 from all parts of the USA, Canada and 18 other nations; the Association strongly encourages graduate students" mem- bership and participation. WHAT DOES IT DO? — Sponsors an annual scientific conference. — Publishes the peer-reviewed Journal of Shellfish Research. — Produces a Quarterly Newsletter. — Interacts with other associations and industry. WHAT CAN IT DO FOR YOU? — You will meet kindred scientists, managers and industry officials at annual meetings. — You will get peer review through presentation of papers at the annual meeting. — If you are young, you will benefit from the experience of your elders. — If you are an elder, you will be rejuvenated by the fresh ideas of youth. — If you are a student, you will make useful contacts for your job search. — If you are a potential employer, you will meet promising young people. — You will receive a scientific journal containing important research articles. — You will receive a Quarterly Newsletter providing information on the Association and its activities, a book review section, information on other societies and their meetings, a job placement section, etc. HOW TO JOIN — Fill out and mail a copy of the application blank below . The dues are 45 US $ per year ($25 for students) and that includes the Journal and the Newsletter! NATIONAL SHELLFISHERIES ASSOCIATION—APPLICATION FOR MEMBERSHIP (NEW MEMBERS ONLY) Name; For the calendar year: Date; Mailinc address: Institutional affiliation, if any: Shellfishery interests: Regular or student membership: Student members only — advisor's signature REQUIRED; Make cheques {MUST be drawn on a US bank), international postal money orders or VISA for S45 ($25 for students with advisor's signature) payable to the National Shellfisheries Association and send to Christine Hodgson. BC Ministry of Ag. and Fisheries, 2500 Cliffe Avenue, Courtenay. British Columbia. CANADA V9R 5M6. INFORMATION FOR CONTRIBUTORS TO THE JOURNAL OF SHELLFISH RESEARCH Original papers dealing with all aspects of shellfish re- search will be considered for publication. Manuscripts will be judged by the editors or other competent reviewers, or both, on the basis of originality, content, merit, clarity of presentation, and interpretations. Each paper should be carefully prepared in the style followed m Volume 13, ( 1 ), of the Journal of Shellfish Research ( 199 1 ) before sub- mission to the Editor. Papers published or to be published in other journals are not acceptable. Title, Short Title, Key Words, and Abstract: The title of the paper should be kept as short as possible. Please include a "short running title" of not more than 48 char- acters including space between words, and approximately seven (7) key words or less. Each manuscript must be ac- companied by a concise, informative abstract, giving the main results of the research reported. The abstract will be published at the beginning of the paper. No separate sum- mary should be included. Text: Manuscripts must be typed double-spaced throughout on one side of the paper, leaving ample margins, with the pages numbered consecutively. Scientific names of species should be underlined and, when first mentioned in the text, should be followed by the authority. Common and scientific names of organisms should be in accordance with American Fisheries Society Special Publications 16 and 17: Common and Scientific Names of Aquatic Invertebrates from the United Slates and Canada: Molhisks and CSNAIUSC: Decapod Crustaceans. Abbreviations, Style, Numbers: Authors should follow the style recommended by the fourth edition (1978) of the Council of Biology Editors fCBE] Style Manual, dis- tributed by the American Institute of Biological Sciences. All linear measurements, weights, and volumes should be given in metric units. Tables: Tables, numbered in Arabic, should be on sepa- rate pages with a concise title at the top. Illustrations: Line drawings should be in black ink and planned so that important details will be clear after reduc- tion to page size or less. No drawing should be so large that it must be reduced to less than one third of its original size. Photographs and line drawings preferably should be pre- pared so they can be reduced to a size no greater than 17.3 cm X 22.7 cm, and should be planned either to occupy the full width of 17.3 cm or the width of one column, 8.4 cm. Photographs should be glossy with good contrast and should be prepared so they can be reproduced without re- duction. Originals of graphic materials (i.e., line drawings) are preferred and will be returned to the author. Each illus- tration should have the author's name, short paper title, and figure number on the back. Figure legends should be typed on separate sheets and numbered in Arabic. No color illustrations will be accepted unless the author is prepared to cover the cost of associated reproduction and printing. References Cited: References should be listed alphabet- ically at the end of the paper. Abbreviations in this section should be those recommended in the American Standard for Periodical Title Abbreviations, available through the American National Standard Institute, 1430 Broadway, New York. NY 10018. For appropriate citation format, see examples at the end of papers in Volume 10, Number I, of the Journal of Shellfish Research or refer to Chapter 3, pages 51-60 of the CBE Style Manual. Page Charges: Authors or their institutions will be charged $65.00 per printed page. If illustrations and/or tables make up more than one third of the total number of pages, there will be a charge of $30.00 for each page of this material (calculated on the actual amount of page space taken up), regardless of the total length of the article. All page charges are subject to change without notice. Proofs: Page proofs are sent to the corresponding author and must be corrected and returned within seven days. Al- terations other than corrections of printer's errors may be charged to the author(s). Reprints: Reprints of published papers are available at cost to the authors. Information regarding ordering reprints will be available from The Sheridan Press at the time of printing. Cover Photographs: Particularly appropriate photo- graphs may be submitted for consideration for use on the cover of the Journal of Shellfish Research. Black and white photographs, if utilized, are printed at no cost. Color illus- trations may also be considered. Corresponding: An original and two copies of each manuscript submitted for publication consideration should be sent to the Editor, Dr. Sandra E. Shumway, Natural Science Division, Southampton College, LIU Southamp- ton, NY 11968, Ph. 516-287-8407, FAX 516-287-8419. Membership information may be obtained from the Trea- surer using the form in the Journal. Institutional subscribers should send requests to: Journal of Shellfish Research, P.O. Box 465, Hanover, PA 17331. JOURNAL OF SHELLFISH RESEARCH Vol. 15, No. 1 APRIL 1996 CONTENTS Preface — William S. Fisher, Editor 3 Foreword — Frank O. Perkins 5 Sammy M. Ray Histoncal perspective on Perkinsus marinus disease of oysters in the Gulf of Mexico 9 Jay D. Andrews History of Perkinsus imiriiuis. a pathogen of oysters in Chesapeake Bay 1950-1984 13 Eugene M. Burreson & Lisa M. Ragone Calvo Epizootiology of Perkinsus marinus disease of oysters in Chesapeake Bay, with emphasis on data since 1985 17 Thomas M. Soniat Epizootiology of Perkinsus marinus disease of eastern oysters in the Gulf of Mexico 35 Susan E. Ford Range extension by the oyster parasite Perkinsus marinus into the northeastern United States: response to climate change? 45 Fu-Lin E. Chu Laboratory investigations of susceptibility, infectivity and transmission of Perkinsus marinus in oysters 57 Frank O. Perkins The structure ot Perkinsus marinus (Mackm, Owen & Collier, 1950) Levine, 1978 with comments on taxonomy and phylogeny of Perkinsus spp 67 Jerome F. La Peyre Propagation and in vitro studies of Perkinsus marinus 89 David Bushek & Standish K. Allen, Jr. Races of Perkinsus marinus 103 William S. Fisher t& Leah M. Oliver A whole-oyster procedure for diagnosis of Perkinsus marinus disease using Ray's fluid thioglycollate culture medium 109 Kennedy T. Paynter The effects of Perkinsus marinus infection on physiological processes in the eastern oyster. Crassoslrea virginica 119 Robert S. Anderson Interactions of Perkinsus marinus with humoral factors and hemocytes of Crassostrea virginica 127 Patrick M. Gaffney & David Bushek Genetic aspects of disease resistance in oysters 135 Eric A'. Powell, John M. Klinck & Eileen E. Hofmann Modeling diseased oyster populations. I!. Triggering mechanisms for Perkinsus marinus epizootics 141 G. E. Krantz & S. J. Jordan Management alternatives for protecting Crassoslrea virginica fisheries in Perkinsus marinus enzootic and epizootic areas 167 Victor S. Kennedy The ecological role of the eastern oyster, Crassoslrea virginica, with remarks on disease 1 77 COVER PHOTO: "Micrograph of a mature trophozoite from in vitro culture of Perkinsus marinus. Photo courtesy of Frank Perkins." JOURNAL OF SHELLFISH RESEARCH VOLUME 15, NUMBER 2 JUNE -1996 The Journal of Shellfish Research (formerly Proceedings of the National Shellfisheries Association) is the official publication of the National Shellfisheries Association Editor Dr. Sandra E. Shumway Natural Science Division Southampton College, LIU Southampton, NY 1 1968 Dr. Standish K. Allen, Jr..(1998) Rutgers University ^skin Laboratory for Shellfish f Research |.0. Box 687 Port Norris, New Jersey 08349 Dr. Peter Beninger (1997) Department of Biology University of Moncton Moncton, New Brunswick Canada ElA 3E9 Dr. Andrew Boghen (1997) Department of Biology University of Moncton Moncton, New Brunswick Canada ElA 3E9 Dr. Neil Bourne (1996) Fisheries and Oceans Pacific Biological Station Nanaimo, British Columbia Canada V9R 5K6 Dr. Andrew Brand (1996) University of Liverpool Marine Biological Station Port Erin, Isle of Man Dr. Eugene Burreson (1997) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Peter Cook (1998) Department of Zoology University of Cape Town Rondebosch 7700 Cape Town, South Africa EDITORIAL BOARD Dr. Simon Cragg (1998) Faculty of Technology Buckinghamshire College of Higher Education Queen Alexandra Road High Wycombe Buckinghamshire HPll 2JZ England, United Kingdom Dr. Leroy Creswell (1997) Harbor Branch Oceanographic Institute US Highway 1 North Fort Pierce, Florida 34946 Dr. Lou D'Abramo(1998) Mississippi State University Dept of Wildlife and Fisheries Box 9690 Mississippi State, Mississippi 39762 Dr. Ralph Elston (1996) Batlelle Northwest Marine Sciences Laboratory 439 West Sequim Bay Road Sequim, Washington 98382 Dr. Susan Ford (1998) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Raymond Grizzle (1997) Randall Environmental Studies Center Taylor University Upland, Indiana 46989 Dr. Robert E. Hillman (1998) Battelle Ocean Sciences New England Marine Research Laboratory Duxbury, Massachusetts 02332 Dr. Mark Luckenbach (1997) Virginia Institute of Marine Science Wachapreague, Virginia 23480 Dr. Bruce MacDonald (1997) Department of Biology University of New Brunswick P.O. Box 5050 Saint John, New Brunswick Canada E2L 4L5 Dr. Roger Mann (1998) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Islay D. Marsden (1996) Department of Zoology Canterbury University Christchurch, New Zealand Dr. Kennedy Paynter (1998) 1200 Zoology Psychology Building College Park, Maryland 20742-4415 Dr. Michael A. Rice (1996) Dept. of Fisheries, Animal & Veterinary Science The University of Rhode Island Kingston, Rhode Island 02881 Dr. Tom Soniat (1998) Biology Department Nicholls State University Thibodaux, Louisiana 70310 Susan Waddy (1997) Biological Station St. Andrews, New Brunswick Canada, EOG 2X0 Mr. Gary Wikfors (1998) NOAA/NMFS Rogers Avenue Milford, Connecticut 06460 Journal of Shellfish Research Volume 15, Number 2 ISSN: 00775711 June 1996 Mwlpl ■^ ■eV^Sqfa^i^ln^'' (iuti(5r\ I "■"fii. 7 IN MEMORIAM R. TUCKER ABBOTT 1919-1995 Figure 1. Photograph of R. Tucker Abbott, circa the early 1970s, measuring gastropod shells at the Delaware Museum of Natural History. Picture a likable, enthusiastic biologist surrounded by an ad- miring group of graduate marine biology students, on the sand flats at low tide near Indian River. DE. Repeatedly plunging his arm deep into siphonal holes in the loose sand, the fit biologist repeatedly comes up with a squirming, squirting razor clam — to the delight of the students who have been digging unsuccessfully for them with shovels and forks, the impact of each thrust merely stimulating the reactive burrowers to bury more deeply! This was Dr. R. Tucker Abbott, a friendly, energetic, schol- arly, highly productive systematic malacologist. He was as much at home in the field collecting and studying mollusks as curating them in museum collections, writing scientific papers and mono- graphs, editing equisitely illustrated books and popular manuals about them, and associating with students and amateur and pro- fessional malacologists alike. Tucker died of pulmonary fibrosis illness on November 3. 1995, at the age of 76, at his home on Sanibel Island. FL. He was Founding-Director of the new Bailey-Matthews Shell Museum. Tucker was survived by his charming wife, Cecelia White, who for many years enthusiastically supported him in his professional malacological activities. Tucker was bom September 28. 1919. in Watertown, MA. Little could he have anticipated the full, fruitful, satisfying pro- fessional life that would unfold in the years ahead. These are some of the highlights of his career: 1938^2. Research Assistant in the Museum of Comparative Zo- ology, Harvard University, MA. 1942. Received a B.S. degree at Harvard College. 1942^4. Was a United States Naval Aviator (Lt , USNR) work- ing as a dive bomber pilot. 1944—46. Malacologist with the United Sta'js Naval Medical Re- search Unit 2 (Lt., USNR); the first medical malacologist m history to attempt to control schistosomiasis, the fata' blood fluke parasite. Studies took hiiii to Guam, Marianas, and fi- nally to China's Yangtze Valley, where he discovered the life cycle of the schistosome in a small freshwater snail. 1946-49. Assistant Curator in the Division of Mollusks. U.S. National Museum, Smithsonian Institution, Washington. DC. 185 186 In Memoriam: R. Tucker Abbott 1949. Ri .ived an M.S. degree from nearby George Washington Uni -rsity, Washington, D.C. 1949-54. Associate Curator in the Division of Mollusks, U.S. National Museum. While there, he prepared the first edition of his American Seashells. 1955. Obtained the Ph.D. degree from George Washington Uni- versity. 1954-69. Held the Pilsbry Chair of Malacology and the Chair- manship of the Department of Mollusks, Academy of Natural Sciences, Philadelphia. During this period, he was actively writing, editing, and publishing (see Partial Bibliography). Many of his works have been translated into several other languages. 1969-76. Held the du Pont Chair of Malacology and Chairman- ship of the Department of Mollusks, as well as the Assistant Directorship of the Delaware Museum of Natural History, Wil- mington. There he continued writing actively and increased the number of speaking engagements, especially to shell collectors and shell clubs, in many regions of the United States. 1972. Became president of his company, American Malacologists, Inc. Publishers of Distinctive Books on Mollusks. 1973. Accepted the active honorary position of Adjunct Professor in the College of Marine Studies, University of Delaware. 1979. Resigned as Adjunct Professor in the College of Marine Studies when he moved to Florida to continue writing, pub- lishing books, and consulting in malacology. There his literary pursuits continued actively, and in addition, he began publish- ing the systematic malacological works of other scholars. At the time, he was also involved with the Conchologists of Amer- ica, but his overriding interest became the development of the Bailey-Matthews Shell Museum on Sanibel Island. Subse- quently he was honored as the Founding Director of the Mu- seum, a facility of the Shell Museum and Education Founda- tion, Inc. Tucker remained active in this capacity until his death. Tucker's research emphasis was distinctly systematic mal- acology, and in this field, he had few peers. This attention was strongly complemented by his intense and rewarding interest in collecting for and building museum collections. His work often took him far afield. The museums that benefitted most from these missions were the Philadelphia Academy of Natural Sciences, the Delaware Museum of Natural History, and the Bailey-Matthews Shell Museum. Tucker listed his principal expeditions as follows (Abbott 1987); 1934—76. Many trips to Bermuda. 1939-40. Harvard-Archbold Expedition to Melanesia and Polyne- sia. 1939, 1944, 1958. Philippines. 1939, 1945. China. 1942, 1944, 1946. Cuba. 1944-45. Marianas. 1952. National Research Council Expedition to East Africa. 1963. Anton Bruun Cruise to Bay of Bengal. 1970. Grand Cayman. 1972. Solomons. 1983. Bahamas, Senegal, Seychelles, Sri Lanka, Thailand, Indo- nesia, Australia, Tasmania, New Zealand, Tahiti. 1984. New Guinea and Admiralty Islands. Tucker's outstanding international reputation is well de- served. In addition to his many other accomplishments, he pub- lished over 200 major and lesser works, of which 14 were major books (see Partial Bibliography); wrote dozens of books reviews; edited and published many books by other writers; and described one new family, 10 new genera or subgenera, and 70 new species of mollusks, many of these probably resulting from new discov- eries during oceanic expeditions. Tucker was widely recognized as an eminent malacological systematist. He received five major awards for his literary efforts between 1953 and 1978, was listed in some 15 . . . Who's Who . . . directories, was associated with some 17 shell clubs, and was a member of malacological societies as far away as Australia and Uruguay. But of the latter, he was probably most committed to the American Malacological Union, serving in several offices and finally as president in 1959. My long friendship with Tucker began in 1954 when we cor- responded on the taxonomy of the large ecologic form of oyster drills {Urosalpiiix cinerea foUyensis] from the Eastern Shore of Maryland and Virginia. He was then with the United States Na- tional Museum. During his subsequent tenures at the Academy of Natural Sciences, at the Delaware Museum of Natural History, and finally, in Melbourne and Sanibel Island, we continued to discuss systematic problems most frequently at meetings of the American Malacological Union. Ours was a long and extremely cordial association, especially during Tucker's stint at the Dela- ware Museum of Natural History, 1969-1977. Understandably, I was delighted by the presence of a close malacological colleague so near at hand. In the fall of 1973, shortly after I joined the College of Marine Studies, University of Delaware, in Lewes, and on my recommen- dation. Dean William Gaither of the College of Marine Studies invited Tucker to join the College of Marine Studies as an Adjunct Professor. He was to co-teach with me a graduate course in mal- acology, instruct a course of his choosing, and serve on graduate committees. Tucker graciously accepted our invitation, and the faculty welcomed him cordially, proud to have him on our faculty. Tucker co-taught a course in malacology with me in the spring of 1974 and the falls of 1975, 1976. 1977, and 1978. Among the graduate students who assisted us in the malacology course were M. G. (Jerry) Harasewych and Robert Prezant. Enrollment in the different classes ranged from a half dozen to 20 students. Classes in the systematic aspects of the malacology course were held in the Mollusk Department of the Delaware Museum of Natural History, about 100 miles north of Lewes. Tucker illus- trated his lectures with molluscan specimens from the impressively large collections of the museum (in excess of one million) and his professional color transparencies of living mollusks and their hab- itats. Classes were informal, attentatively relaxed, and often punc- tuated by animated discussions. During lunch we gathered around a large table in free space surrounded by some 500 museum shell cabinets, munched sandwiches, related experiences, and enjoyed hearing about Tucker's lively shell-collecting adventures around the world. Students were outspokenly impressed by the diversity of the specimens displayed, the wide range of habitats and geo- graphic regions represented, and the intricacy of the nomenclature and classification of some of the taxa. The remainder of the malacology course was held at the Col- lege of Marine Studies on the Delaware coast in Lewes. Ecologic field trips were taken by car to representative sand and mud flats, ocean beaches, and salt marshes and by boat on Delaware Bay. These trips, taken early during each course, not only "broke the ice" between students and professors, but they also provided ex- periences in the earthy exercise of slopping over mud flats and marshes, collecting live marsh snails, burrowing bivalves from sand, mud, or peat substrata, and dredging mollusks from the In Memoriam: R. Tucker Abbott 187 bottom of the Delaware Bay. Even on cool, rainy days, students bantered among themselves as they pulled an occasional class member out of a marsh ditch or hole or prevented another from sliding overboard in a rough sea. Collected live animals were maintained in running seawatcr in the Lewes laboratory for later functional studies. Using the systematic parts of the course as a foundation, we concentrated on the anatomical, behavioral, and functional biol- ogy of live mollusks in representative local taxa. Teachers, the teaching assistant, and students all participated in the lectures, discussions, and laboratory work. To all of this Tucker contributed substantially, not only by his fine reputation as a scientist, but also by his broad knowledge, experience, wit, warmth, and empathy for the students. Students still comment to me how much they appreciate the opportunity of having studied with Tucker. My gain likewise has been great. The favorite laboratory for students was a SEM study of a hard part of a mollusk of their choice. For his own course. Tucker chose to teach a Winterim Course in Evolutionary Biology on the main campus of the University of Delaware in Newark. The course was taught dunng the years 1975, 1976, and 1977, with the same success that greeted Tucker during the instruction of our joint course in malacology. Tucker was especially helpful to graduate students while serv- ing on their graduate committees. He made available the very large molluscan collections of the Delaware Museum of Natural History and provided especially helpful advice on biosystematics and its application in thesis and dissertation research. Students on whose graduate committees Tucker served were Margaret Carter. Clem- ent Counts II. G. M. (Jerry) Harasewych. Peter Kinner. and Rob- ert Prezant. He was quick to discuss molluscan systematics with other students and faculty who sought his advice. Reminiscing, Clem Counts once noted that Tucker was fond of autographing copies of his many books on mollusks. This practice was well known to his peers, and Tucker himself laughed about his "inability to refuse to sign" a cover page. At the 1990 meeting of the American Malacological Union in Woods Hole, MA, Clem was standing in the back of the lecture hall with Tucker during the book and shell auction. After a tmie, a copy of Tucker's classic American Seashells was brought out. The copy, coming directly from the publisher, was wrapped in plastic, which prompted Rich- ard Petit, the auctioneer, to quip that the wrapper was significant because it indicated that the book was "untouched by the pen of R. Tucker Abbott." Chuckling. Tucker turned to Clem and counter- quipped that "The book should fetch a very high price sine? there were fewer unsigned, than autographed copies!" In the summer of 1979. Tucker resigned as Adjunct Professor in the College of Marine Studies in preparation for his permanent move to Florida. He had terminated his position with the Delaware Museum of Natural History late in 1977. On Tucker's departure. Dean William Gaither. College of Marine Studies, and I wrote Tucker expressing our deep regret for his departure, but gratitude for his many contributions to the University of Delaware during his tenure as Adjunct Professor. Going to Florida pennitted Tucker to concentrate on his writing and to continue building his successful publishing company. American Malacologists. Inc. Now his earlier wish to more seri- ously pursue writing could be fully realized. Tucker spent the remainder of his life collecting and studying mollusks. writing, and publishing, among other literature, a remarkable series of enthusiastically received books for amateur naturalists and collec- tors. The formal opening of the Bailey-Matthews Shell Museum took place on November 18. 1995. just 2 weeks after Tucker's death. The museum was the materialization of his vision of a "monument to shells for people, not just a museum full of shells" (Scheu 1995). Although Tucker did not live to see the formal opening of the museum, he did take pleasure in seeing paying visitors pass through the halls of the museum earlier that year. "On November 3. 1995, the world lost an extraordinarily re- spected man of science and a godfather to shellers everywhere!" (Hallstead 1995); and I would add, an extraordinarily gifted man of letters who had the natural talent of bridging between amateur and professional malacologists to the benefit of both as well as to the benefit of the field of malacology. ACKNOWLEDGMENTS I am indebted to Jerry Harasewych for a list of Tucker's pub- lications, from which the following "Partial List of Publications" was taken; and to Cecelia Abbott, Russell Jensen, and Paula Mik- kelsen for background information. The photograph of Tucker was kindly provided by the Delaware Museum of Natural History, courtesy of Paula Mikkelsen. CITATIONS AbboU. R, T. 1987. Living American Malacologists and Private Shell Collectors. 1986-1987. pp. 1-2. American Malacologists. Inc., Mel- bourne, Florida. Hallstead, B. 1995. Dr. R. Tucker Abbott 1919-1995. p. 3. Newsletter. Bailey-Matthews Shell Museum. Inc.. Sanibel Island. Florida. Scheu, L. 1995. Robert Tucker Abbott. Am. Conchologists 23(4):3-10. PARTIAL. CHRONOLOGICAL LIST OF Clench. W. J. & R. T. Abbott, 1941 . The genus Strombi(s in the Western Atlantic. Johnsonia 1(1); 1-15. Clench, W. J. & R. T, Abbott. 1942. The genera Tectarius and Echinimus in the Western Atlantic. Johnsonia 1(4): 1—4. Clench. W. J. & R. T. Abbott. 1943. The genera. Cypraeacassis. Monim. Sconsia and Dalium in the Western Atlantic. Johnsonia 1(9): 1-8. Clench W. J. & R. T. Abbott. 1943. The genera Gaza and Livona in the Western Atlantic. Johnsonia I(12);l-9. AbboU. R. T. 1943. Guantanamo Bay. Cuba Jolmsoma 1(12):10-1L Abbott. R. T. 1944. The genus Modulus in the Western Atlantic, Jolmso- nia 1(14): 1-6. Abbott. R. T. 1945. A new Celebes freshwater snail (Hydrobiidae). Occ. Papers Mollusks. Hunurd 1(1): 1-4. PUBLICATIONS BY R. TUCKER ABBOTT Clench, W. J, & R. T. AbboU. 1945. The genus Siromhus in ''.e Western Atlantic. Johnsonia 1(18):1. Abbott. R. T. 1945. The Philippine intermediate sn;M host (Schislo- mophora quadrasi) of schistosomiasis. Occ. Pa' ers Mollusks. Har- vard 1(2):5-16. Abbott. Lt. R. T. 1946. The egg and breedi'^.g habits of Oncomelania quadrasi Mlldff.. the schistosomiasis s-.ail of the Philippines I'cc. Papers Mollusks. Hanard 1(6):41-4S;. Abbott, R. T. 1948. Handbook of Medically Important Mollusks of the Onent and Western Pacific. Bull. Mus. Comp. Zool., Harvard 100(3): 243-328, pis. 1-5. Abbott. R. T. 1948. Mollusks and medicine in World War II. Repi. Smith- sonian Ins! . 1947:325-338. In Memoriam: R. Tucker Abbott Abbott. R 1 . 1948. A new genus and species of Philippine Amnicolidae. Nam ui61(3);75-80. pi. 5. Jaume, M. L. & R. T. Abbott. 1948. A new Cuban species of the Am- Tucolid genus Nanivitrea. Revista Soc. Matacologka "Carlos n(m was responsible for Scotian estuaries. Ship Harbour, again developed red digestive the red discolouration of oyster digestive glands in Lakj Grevelin- glands. Preserved phytoplankton samples obtained from local gen in the Netherlands. mussel growers as part of the Phytoplankton Monitoring Program The following study was undertaken to clarif, the relationship (Carver et al. 1992) were found to contain high numbers of 30- to between the occurrence of red digestive gland in cultured mussels 60-|ji.m particles. Although difficult to identify initially, they were and the abundance of A-/, rubrum. The spc .fie objectives were (ai found to exhibit the same phycoerythrin-like fluorescence as tissue to demonstrate that the discolouration ■ . the digestive gland ,vas samples from the digestive gland. Subsequent examination of un- due to the presence of phycoerythrir., (b) to assess the rate of the preserved water samples revealed high concentrations of the pho- uptake and depuration of phycoerythrin by the mussels: (c) to tosynthetic ciliate Mesoduiium rubrum. This delicate species var- determine whether cultured sea scallops would also accumulate ies in size from 20 to 70 iJim and harbours an algal (crvptomonadi phycoerythrin; and (d) to document the temporal and spatial dis- endosymbiont containing the accessory photosynthetic pigment tribution of A/, rubrum in Ship Harbour. 191 192 Carver et al. MATERIALS AND METHODS Phycoerythrin Fluorescence Extracts of digestive gland from cultured mussels grown m Ship Harbour (Fig. 1) and a control site near Lunenburg, N.S.. were obtained by gently teasing the tissue in seawater. The result- ing slurry was filtered onto 25-mm GF/F filters and examined by spectrofluorometry. Fluorescence (excitation and emission) spec- tra were acquired on a SPEX Fluorolog Fl 11 A spectrofluorometer interfaced with a SPEX DM3000 IBM-compatible computer. Ex- citation and emission wavelengths were scanned and recorded at 0.5-nm intervals. Uptake of Phycoerythrin Ship Harbour is a long, narrow estuary (8 x 1 km) with the deepest water located at the head of the system (Fig. 1). Three mussel longlines were selected as experimental sites: Location 1 (13-m deep) at the upper end, near the river mouth; Location 2 (11-m deep) roughly 2 km downstream: and Location 3 (6-m deep), another 2 km downstream. Preliminary sampling trips were undertaken in April 1992, and a weekly monitoring program was conducted from May to mid- August 1992. To document the uptake of phycoerythrin, "control" mussels (i.e. , with no evidence of phycoerythrin) were transferred from the Lunenburg site into Ship Harbour at weekly intervals (April 10 to June 5). Twenty mussels were placed In each of six Japanese pearl nets, which were then deployed at Location I (3 and 8 ni). Loca- tion 2 (3 and 8 m), and Location 3 (3 and 5 m). After 1 wk, the digestive glands of the control mussels were examined for signs of red colouration or macroscopic evidence of phycoerythrin accu- mulation. This qualitative visual assessment was based on several criteria: (a) colour of the tissue; (b) size of the gland: (c) colour of the crystalline style; and (d) presence of red fluid in the style sac. A mussel with a red and swollen digestive gland, a pink-stained style, and red fluid in the stomach was assigned a rank of "4"; with no red fluid, a rank of "3"; with a clear style, a rank of "2"; with no swelling of the digestive gland, a rank of "1": and with no red colouration (i.e., "normal"), a rank of "0". The rankings were then averaged to obtain an overall value or "red colouration index" for each location. Ship Harbour mussel sleeves, originally hanging from the longline at each location, were also sampled weekly and ranked on the basis of the same criteria. On several occasions, the control mussels held for 1 wk in a pearl net at Location I (3 m) were sampled for histological as- sessment of phycoerythrin levels. These individuals were fixed in 1% glutaraldehyde: 4% formalin, embedded in paraffin, and sec- tioned (6 jxm) through the digestive gland. Unstained tissue sec- tions were examined under epifluorescent illumination (545 nm) with a Reichart-Jung Polyvar 1 microscope. An image analysis system equipped with a black-and-white camera was used to assess the intensity of the fluorescence, which was then used to derive a relative phycoerythrin ranking for each mussel. Values ranged from a maximum of "100%" down to "0%" for the control mussels before being transferred into Ship Harbour. Figure 1 . Map of the Ship Harbour estuary indicating the position of the three sampling locations (LI, L2, and L3), Note that the deepest section of the estuary (14 m) is located at the upper end. Red-Coloured Digestive Glands in Mussels 193 A short-term field-grazing experiment was conducted on May 1. In this case, control mussels were deployed at Location 1. and samples for the histological assessment of phycoerythrin were taken hourly over the next 4 h. To determine whether other species would also develop red digestive glands, "control"" scallops were obtained on three oc- casions from a scallop culture site in Mahone Bay (Fig. 1). The scallops were deployed in pearl nets (six per net) adjacent lo the control mussels at the three locations in Ship Harbour. After 1 wk, the scallops were examined for signs of red colouration and two were processed for histological assessment of phycoerythrin lev- els. Depuration of Phycoerythrin Depuration experiments were carried out on three occasions using mussels with a strong red colouration in their digestive glands. On April 10, 1992, 80 mussels from Ship Harbour were transferred to two tanks of filtered sea water (40 mussels/tank) at the Halifax Fisheries Laboratory (Department of Fisheries and Oceans. Canada), one set at ambient temperature (ca. 5°C) and the other at 15°C. A second set of tanks was stocked with 80 control mussels originally from the Lunenburg site. Ten mussels were sampled from each tank at 20, 53, and 77 h; five were visually assessed for red colouration, and five were prepared for histolog- ical examination. Two similar depuration trials were set up on April 20 and 24 with new groups of Ship Harbour and Lunenburg mussels. A longer term depuration experiment was initiated on May 26; in this case, mussels from Ship Harbour were left in flowing sand-filtered seawater for 5 wk. Seven mussels were sampled ini- tially and after 1, 2, 3, and 5 wk; three were examined for red colouration, and four were prepared for histological assessment of phycoerythrin levels. Abundance and Distribution ofM. rubrum Preliminary phytoplankton samples collected during April 1992 indicated that M. nihniiii was the most likely source of the phycoerythrin accumulated by the mussels. A weekly sampling program was therefore undertaken from May to mid-August to document the abundance and distribution of this species. On each sampling day, temperature and salinity profiles were obtained with a Conductivity-Temperature-Depth meter, and a vertically inte- grated phytoplankton sample was collected with a 20-|xm-pore- size mesh net. Integrated water column samples (0-3, 0-8, and 0-13 m at Location 1 , 0-3 and 0-1 1 m at Location 2. and 0-6 m at Location 3) were collected with a 2-cm-diameter PVC hose. Each sample was drained into a bucket and stirred gently, and a 1,000-ml subsample was removed. Three methods of preserving M. rubrum were initially compared; (a) 1% glutaraldehyde, (b) 0.5% Lugol's iodine ( 100 g of KI, 50 g of iodine, and 100 ml of glacial acetic acid in 1 1 of distilled water), and (c) 2% formalin; acetic acid (50;50). Three subsamples of 30 ml were taken from each sample, filtered onto l-|j.m-pore-size Nuclepore filters, and transferred to slides by freezing (Hewes and Holm-Hansen 1983, adapted by K. Pauley, unpubl. manuscript). Cells of M. rubrum were enumerated at 100 x magnification. Estimates of A/, rubrum concentration in the 3- to 8-m or 8- to 13-m zone were calculated by difference; for example, the value for 3-8 was obtained by subtracting the 0-3 m value from the Q-i ni estimate. RESULTS Phycoerythrin in Mussel Digestive Gland Tissue Spectrofluorometric examination of samples of red-coloured digestive gland indicated the presence of the accessory photosyn- thetic pigment phycoerythrin. Excitation spectra of the chlorophyll tluorescencc of digestive gland slurries possessed a principal peak at 545 nm with secondary peaks at 438 and 675 nm (Fig. 2a). The peak at 545 nm is characteristic of phycoerythrin absorption. 05 Q. U_ 0} o c cu o 05 05 > cr 400 450 500 550 600 650 700 Wavelength (nm) 1 4- b Emission X = 576nm /""^N / \ 0) o O (D~ 05 o o — _5 3- 2- / > CO 1 - J cr y^ 350 400 450 500 Wavelength (nm) 550 05 Q. 0) o c 05 o o > DC 5 - c A Excitation X = 545 nm 4 - 3 - 2 - 1 - 1 1 1 550 600 650 Wavelength (nm) 700 750 Figure 2. Fluorescence excitatiun and emission sp otra of seawater extracts containing red-coloured pigment obtain j from the digestive glands of mussels: (a) excitation spectrum of e iract with fluorescence monitored at 730 nm, the wavelength cho^ n to detect chlorophyll a emission; (b) excitation spectrum of extract with fluorescence emission monitored at 576 nm, the emission maximum of phycoerythrin; the broad peak centered at 545 nm is characteristic of phycoerythrin ab- sorption; (c) emission spectrum of extract using an excitation of 545 nm corresponding to the peak absorption of phycoerythrin: the emis- sion peak at 576 nm is characteristic of phycoerythrin. cps = counts per second. 194 Carver et al. wherea- ,\e peaks at 438 and 675 nm are associated with chloro- phyll jsorption. The overall shape of this excitation spectrum is typical of phycoerythrin-bearing cryptophytes and closely resem- bled excitation spectra from cultures of the cryptophyte Rhodomo- nas salina. A fluorescence excitation scan of the red pigment confirmed an absorption maxima at 545 nm (Fig. 2b). whereas a third scan using an excitation wavelength of 545 nm showed a strong emission peak at 576 nm (Fig. 2c). This fluorescence ex- citation-emission signature was consistent with that of the acces- sory pigment phycoerythrin. Spectrofluorometric spectra of diges- tive gland tissue from the control mussels obtained from the Lunenburg site showed no evidence of phycoerythrin. TABLE 1. Mean rankings of red colouration based on the visual assessment of indigenous Ship Harbour mussels (shallow) and Lunenburg control mussels after I wk in Ship Harbour (shallow and deep). Mussel Source Location I Location 2 Location 3 Ship Harbour (bliallow) 3.09 1.15 1.28 Control (shallow) 2.05 0.79 0.68 Control (deep) 1.55 1.85 0.85 Spatial and Temporal Variations in Red Colouration During April and early May. mussel digestive glands were extremely swollen and had a distinctive russet-red colour. When the digestive gland was dissected, the stomach was found to con- tain a pink-stained crystalline style, as well as an abundance of red fluid. Visual examination of Ship Harbour mussels, as well as control mussels deployed for at least 1 wk in Ship Harbour, indi- cated considerable spatial variability in the extent of these char- acteristics (Fig. 3). Over the 8-wk period, the average ranking for the Ship Harbour mussels (shallow) was higher at Location I than at Locations 2 and 3 (Table 1 ). The control mussels deployed near the surface exhibited a similar pattern, although the mean values were consistently lower. A substantial variation in red colouration over depth was observed at Location 2. where the control mussels deployed in the deeper water had the higher rankme. I Ship-Shallow D Control-Shallow Control-Deep 1 Location 1 I IL Location 2 El ["Pi w-i i> Location 3 1 m. A10 A20 ^24 M20 M26 JOS 1 M06 ^ 1992 Figure 3. Red colouration r:mking based on qualitative visual assess- ment of Ship Harbour mus.sels and Lunenburg control mussels after 1 wk in Ship Harbour. The shallow samples were taken from approxi- mately 3 m below the surface, whereas the deep samples were from 8 m below the surface at Locations 1 and 2 and 5 m below the surface at Location 3. Temporal variation in red colouration was also evident over the 8-wk period from April 10 to June 5 (Fig. 3). In April, mussels from Location 1 (shallow and deep) in the upper estuary and those deployed at depth at Location 2 showed the most intense coloura- tion. In early May. there was generally an increase in the level of red colouration at all depths and locations. After May 15. the control mussels showed little evidence of phycoerythrin accumu- lation, whereas the Ship Harbour mussels continued to exhibit red digestive glands for several weeks. During the summer, mussels sampled from 8 m at Location 1 occasionally showed signs of red colouration, but samples from 3 m and from the other two loca- tions consistently appeared normal. Histological Assessment of Phycoerythrin Histological sections of the red-coloured digestive glands of the Ship Harbour mussels were markedly different from those of con- trol mussels when they were initially obtained from Lunenburg. In particular, they appeared to have larger digestive tubules and less space between the tubules. Particles of red-fluorescing pigment, probably cell fragments of M. ruhrum. were often observed in the stomach area. In the short-term grazing experiment (May 1 ). con- trol mussels showed signs of red-stained styles and red fluid in the style sac within 3 h of deployment. Phycoerythrin levels in the digestive tubules were assessed at 10% after 3 h and 20% after 4 h. indicating rapid accumulation of pigment. This was consistent with the observation that control mussels transferred into Ship Harbour typically resembled the adjacent Ship Harbour mussels within a week of deployment (Fig. 3; April 20 to May 15). Ship Harbour mussels sampled from Location 1 (shallow) be- tween April 10 and May 15 consistently had phycoerythrin levels of 90-100% (Fig. 4) By early June, pigment levels had declined to 80% , and there was no visual evidence of red colouration in the digestive gland. From mid-June to mid-July, phycoerythrin levels ranged from 27 to 56%. with considerable variation among indi- viduals, possibly because of differences in depuration rates. Interestingly, control scallops transferred into Ship Harbour (April 24 and May 1) with the control mussels showed no visual signs of red colouration after 1 wk. Examination of the digestive gland tissues under epifluorescence. however, indicated the pres- ence of phycoerythrin. Levels were estimated at 60-70%, as op- posed to the adjacent control mussels at 100%. At first, these results suggested that scallops might not accumulate sufficient phycoerythrin to display signs of red colouration. However, scal- lops from another nearby site (Country Harbour) that had higher concentrations of M. riibrum (80.000 cells • 1"') did show red- coloured digestive glands (Fig. 5a). Red-Coloured Digestive Glands in Mussels 195 Figure 5. (a) Example of a cultured sea scalliip with red-coloured digestive gland from Country Harbour, an estuarj located KM) km east ■ ship Harbour (concentration of A/, ruhriim. 80,000 cells l~'); note that control scallops deployed in Ship Harbor accumulated phycoerythrin but not to the extent that they developed red colouration; scale bar = 15 mm; (bl M. rubriim under differential interference contrast, scale bar = 15 Jim; (c) M. ruhriim under green epifluorescent illumination showing the characteristic orange-red autofluorescence of the phycoerythrin pigment in the chloroplasts, scale bar = 15 p.m; (d-g) cross-sections of mussel digestive gland illustrating the depuration of phycoerythrin over 3 wk; the relative phycoerythrin ranking declined from an initial value of 100% to 80% (Week 1 ), 40% (Week 2). and 10% (Week 3); scale bar = 100 pim. 196 Carver et al. e 80 20 "-^'-t-y i ^ k - ^ k : V- 1 . 1 , 1 May June 1992 July Figure 4. Phycoerv thrin fluorescence index ( '^c ) for the digestive gland of Ship Harbour mussels from Location 1 (3 m). Depuration Experiments The depuration experiments were undertaken in an attempt to document the rate at which the red colouration disappeared from the digestive gland of the mussels. Visual assessment over time (Fig. 6) indicated a rapid change in the extent of the colouration; in particular, the red fluid and the pink colour of the style virtually disappeared after 20 h. The digestive gland, however, remained abnormally large relative to that of the control mussels; a notice- able decrease in size was only observed at 77 h. Assessment of phycoerythrin using epifluorescence indicated a slight decline over the duration of the experiment, but the fluorescence emission was still strong (i.e., 7(3-80'7<:) after 77 h. Phycoerythrin apparently remams in the digestive gland long after visual evidence of pig- ment accumulation disappears. There was no significant difference in the rate of depuration between mussels in the ambient tank at 5°C and those in the 15°C tank. In the long-term depuration experiment (Fig. 5d-g, Fig. 7), macroscopic evidence of phycoerythrin disappeared within 1 wk, but examination of mussel tissues under epifluorescent illumina- tion indicated that phycoerythrin persisted in the digestive gland for approximately 3 wk. Depuration rates calculated from this trial suggested a decline in phycoerythrin levels of approximately 22% per day (r = 0.92). Temperature and Salinity Conditions in Ship Harbour Preliminary measurements during April indicated a mean water column temperature of 3-5°C in Ship Harbour. Temperatures in- Figure 7. Phycoerythrin fluorescence index (%) for the digestive gland of mussels depurated in filtered seawater over 5 wk. See also Figure 5d-g. creased sharply from 4 to 6°C in eariy May (Fig. 8) to >10°C in early June. In general, the mean temperature at the deepest site (Location 1 ) near the head of the estuary was lower than that at the two more oceanic but shallower sites (Locations 2 and 3). Contour diagrams of temperature vs. depth over time and salinity vs. depth over time are presented for Location 1 (Fig. 9). Following the formation of a vertically stratified water column in early June, temperature in the 8- to 13-m zone remained below 10°C through- out the summer. Location 2 also showed <10°C temperatures at 8-10 m. but this cold layer was not evident at the more shallow Location 3. The salinity contour diagram indicated a warm low- salinity layer (>10°C. <28°/oo) in the upper 2 m from mid-May to mid-June, followed by a mixing event and the reestablishment of a stable halocline in late June. Profiles for Locations 2 and 3 showed a similar vertical structure, although the low-salinity sur- face layer was less evident towards the mouth of the estuary. Abundance and Distribution o/M. rubrum Initial identification of the photosynthetic ciliate M. rubrum was based primarily on the presence of a distinct row of cilia at the juncture of the two body sections (Fig. 5b) and observations of phycoerythrin-containing chloroplasts under epifluorescent illumi- nation (Fig. 5c). Initial attempts to fix M. rubrum with glutaral- dehyde proved unsuccessful, and although Lugol's iodine was found to be quite effective in preservmg the delicate cilia of this species, the best results were obtained with 2% formaliniacetic Days Figure 6. Decline in the red colouration of the digestive gland of mus- Figure 8. Mean water column temperature at the three locations in sels depurated in filtered seawater for 77 h. Ship Harbour. Red-Coloured Digestive Glands in Mussels 197 26 J03 1992 Figure 9. Contour plots of (a) the temperature structure and (bl the salinity structure of the water column at Location 1. acid (50:50). Preservation of samples immediately after collection was also found to reduce the extent of cell disruption and minimize damage to the cilia. Weekly estimates of M. rubrum concentration showed substan- tial temporal and spatial variability (Fig. 10). It should be noted that prelimmary samples collected in April ranged from 5.000 to 50.000 cells -1"'. In early May. the abundance of M. rubrum was relatively low (6,500-13.000 cells • P'). but numbers in- creased again in the week of May 6 ( 17.000-26,000 cells • 1~ '). By May 15. M. rubrum was on the decline while other algal species were increasing in diversity and abundance. This change In the phytoplankton assemblage coincided with a decline in the phy- coerythnn levels in the digestive glands of the mussels (Fig. 4). M. 25,000 . P \ L1 •■-■-- L2 L3 * * 15 000 10,000 5.000 \ w •? ..... ^ / \ / -ir \ /...x. 1- ■■'■■< — >"\\, rubrum remained below 6.000 cells • I " ' through the summer with the exception of a brief bloom at Location 1 in late July. Depth profile sampling indicated that M. rubrum was not evenly distributed through the water column; for example, on May 1. the concentration in the upper 3 m (3.700 cells • P') was substantially lower than the average value for the water column (6.500 cells ■ I" '), thus indicating that the species was relatively more abundant below 3 m. Further samples collected during June and July (Fig. 11) suggested that, as temperature increased, the population became confined to the deeper zones of the water col- umn. By early July, most of the M. rubrum population was found in the 8- to 13-m zone at Location 1 or the 3- to I l-m zone at Location 2. Temporal variability in the vertical distnbution of M. rubrum may have been due to the horizontal movement of water masses of varying cell abundance or to the actual vertical movement of cells in response to light and/or tidal currents. For example, on May 6. water samples collected only I h apart showed substantial changes in cell distribution. To investigate short-term changes in the ver- tical distribution of M. rubrum. samples were taken every 2 h from 1 130 to 2030 h at Locations 1 and 2 on May 20 (Fig. 12). Abun- dance estimates for the whole water column varied from 8.700 to 1 1 .900 cells ■ r ' at Location I and 9. 100 to 16.500 cells • r ' at Location 2. The vertical distribution of M. rubrum varied substan- tially over the 9-h period, but there was little evidence of a con- sistent tidal or diurnal migration pattern. At Location 1 during the ebb tide, there appeared to be a net movement of cells out of the surface layer (0-3 m) and the bottom layer (8-13 m) into the midwater zone (3-8 m). At Location 2. there appeared to be a 15.000 5.000 Ma June July August 1992 Figure 10. Abundance of M. rubrum (cells 1~') averaged over the whole water column at each location from May 1 to August 18. 5.000 M26 JOS J12 J18 J26 JOl J'j 1992 Figure 11. Vertical distribution o( M. rubrum at Locations 1 and 2. Note that there were no stratiPied samples from June 5 and 12, and therefore, the single estimate from the whole water column was used for each layer. 198 Carver et al. □ 0-3 m £-3 3-8 m ■ 8-14 m j Location 1 High Tide ^Ebb »- Low Tide ^ Flood 11 35 17:40 11 30 13 30 16 00 18 30 20 30 May 2 1992 Figure 12. Vertical distribution of M. rubriim over a tidal cycle at Locations 1 and 2. decline in the concentration of cells near the surface (()-3 m) during the ebb tide, but there was no consistent temporal pattern. DISCUSSION Accumulation and Depuration of Phycoerythrin Spectrofluorometry was used to demonstrate that the red co- louration in the mussel digestive gland was due to the presence of the accessory photosynthetic pigment phycoerythrin. The shape and the peaks of the chlorophyll excitation spectrum were typical of cryptophytes. or ciliates such as M. ruhruin. which possess cryptophyte-typc chloroplasts. Subsequent surveys of the phyto- plankton assemblage using epifluorescence microscopy revealed that the most likely source of this red pigment was the photosyn- thetic ciliate M. nibnim. Although this study was initiated after the start of the M. ruhrum bloom in mid-March (Phytoplankton Mon- itoring, unpubl. data), it did serve to clarify the temporal relation- ship between the presence of this species and the occurrence of red-coloured digestive glands in the mussels. In particular, the decline in the abundance of M. ruhrum in the last 2 wk of May coincided wu'i the disappearance of the red colouration and a decrease in phy^ erythrin levels in the digestive gland. A concur- rent increase in the availability of other nonphycoerythrin food Nources may have contributed to the observed decline in pigment levels. As expected, visual assessment of phycoerythrin levels was much less sensitive than microscopic evaluation under epifluores- cent illumination. Only when phycoerythrin levels exceeded 80% was there macroscopic evidence of pigment accumulation in the digestive gland. For example, the apparent lack of red colouration in the digestive glands of the sea scallops initially suggested that this species was not feeding on M. ruhrum. whereas in actuality, phycoerythrin levels were simply below the visual detection limit. Differences in the rate of uptake of phycoerythrin between scallops and mussels may indicate that scallops can select against M. ru- hrum. or that they are more effective at metabolizing and/or ex- creting the phycoerythrin pigment. Control mussels deployed in Ship Harbour after May 15 rarely accumulated sufficient phycoerythrin to show signs of red coloura- tion. The indigenous Ship Harbour mussels, however, continued to appear red for 2-3 wk, suggesting that depuration rates only slightly exceeded uptake rates. Laboratory depuration trials indi- cated that, even in cases where mussels exhibited strong coloura- tion, these characteristics disappeared very rapidly in sand-filtered seawater. In particular, the red fluid from the style sac as well as the pink colouration of the style disappeared within 48 h. Full depuration of phycoerythrin may require a further 4-5 wk, but it appears that mussels could be marketed during this period. Thus, in some instances, short-term depuration of mussels with red- coloured digestive glands could prove a viable option for the shell- fish industry. It remains to be determined whether phycoerythrin. which ac- cumulates in the digestive gland, is metabolized or excreted by mussels and/or scallops. Several researchers have used the auto- fluorescence of photosynthetic pigments and their breakdown products to investigate the nutritional history of bivalves (Bncclj and Malouf 1984, Hummel 1985, Robinson et al. 1989, Smith and Wikfors 1992). Because phycoerythrin is readily detectable by spectrofluorometry or epifluorescence microscopy, it may have some utility as a natural tracer for feeding, digestion, and/or en- ergy storage studies. For example, the ratio of phycoerythrin to chlorophyll in the stomach relative to that in the phytoplankton may indicate whether bivalve shellfish are selecting for, or against. A/, ruhrum. Also, the rate at which phycoerythrin accu- mulates in, or disappears from, the digestive gland may provide insight into basic metabolic processes and how these differ among bivalve species. Occurrence ofM. ruhrum in Ship Harbour The high abundance of M. ruhrum in Ship Harbour was con- sistent with the observation that this species tends to establish substantial populations at the head of estuaries (Crawford 1989). Crawford and Purdie (1992) suggested that M. ruhrum actively disperses away from turbulent surface water on the ebb tide in order to avoid being flushed out of the estuary. The combination of this negative response to turbulence superimposed on a diurnal vertical migration pattern (Smith and Barber 1979, Crawford 1989) may account for some of the short-term variability in the distribution of M. ruhrum in Ship Harbour. The complexity of this behaviour may also have confounded our attempt to identify tidal or diurnal migration patterns. Although M. ruhrum is reported to be positively phototactic (Crawford 1989), stratified sampling consistently indicated higher cell densities in the deeper zones of the water column. These depths were well below the preferred light intensity of this species, i.e., 50% surface illumination (cited in Smith and Barber 1979). As the surface water warmed, this pattern was accentuated; by late June, the only significant population of M. ruhrum remained in the Red-Coloured Digestive Glands in Mussels 199 8- to I3-m zone at Location 1. There are several other reports of M. nihrum populations existing under low-temperature and low- light conditions, e.g., under the permanent ice in Antarctica (Sa- toh and Watanabe 1991 ), deep in the aphotic zone in the Baltic Sea (Leppanen and Bruun 1986), and at the interface of the anoxic layer in a Finnish fjord (Lindholm and Mork 1990), It is probable that the ability of Af . ruhnim to vertically migrate allows it to seek out the optimal combination of light and nutrient conditions. This behavioural advantage may in turn explain the ability of M. ni- hrum to bloom in the early spring, 1-2 mo ahead of other pho- totrophic species in temperate estuaries (Revelante and Gilmartin 1987). An increase in the abundance of M. rubnon was observed in the 3- to 8-m zone at Location 1 from July 24 to 30. During this period, however, it was noted that the cells of M. nibntm were 20-40 |j.m in diameter, as opposed to 30-80 (jim dunng the spring. Montagnes and Lynn ( 1989) observed a similar seasonal variation in the size of Mesodinium in the Gulf of Maine. They suggested that smaller cells, with their higher surface-to-volume ratio, may have an advantage under nutrient-depleted summer conditions. On the other hand. Revelante and Gilmartin ( 1987) implied that there are two forms of Mesodiiuum. the phototrophic Mesodiimim iii- bniin and a smaller, nanoplanktonic (<20-|a.m) ciliate that they labelled Mesodinium sp. In their study of the Damariscotta Estu- ary, the larger form dominated from December through April, whereas the smaller form prevailed from April through July. Lind- holm and Mork ( 1990) also observed a wide range o( Mesadiiuiim phenotypes (20-60 |j.m) in a Baltic fjord; large forms ( >4() p.ni) dominated in some years, and small forms (<30 ixm) did so in other years. An interesting discrepancy was observed between the spring plankton community at Ship Harbour and sites further south along the Atlantic Coast of Nova Scotia. In mid-April, the more south- erly sites exhibited a classic spring bloom with a high abundance and diversity of algal species, whereas Ship Harbour continued to have very low levels of phytoplankton but high numbers of M. rubnon and other ciliated protozoans such as Sfrombidium. Tin- linnopsis, and Didinium. Similar microzooplankton assemblages were observed in the spring in Flodevigen Bay in southern Norway (Dale and Dahl 1987) and in the Damariscotta Estuary in Maine (Revelante and Gilmartin 1987). One possibility is that the nutrient regime in Ship Harbour, particularly at the upper end where the rivers enter, favours the development of M. ruhrum. Nutrient mea- surements from the spring period (P. Strain. Bedford Institute of Oceanography, unpubl. data) indicated unusually low nitrate lev- els, which translated into relatively high phosphate-to-nitrate ra- tios. Interestingly. Fonds and Eisma (1967) linked phosphate-rich upwelled waters with the occurrence of M. rubrum blooms in Holland. Geographic Distribution ofM. rubrum Examination of mussels from several other aquaculture sites along the Atlantic Coast of Nova Scotia revealed no prominent cases of red colouration, with the exception of Country Harbour, where M. rubrum reached concentrations of 80.000 cells ■ 1 " ' in May-June 1992 (Carver, unpubl. data). Most sites showed rela- tively low levels of M. rubrum (500-5.000 cells • P'). although in a few instances, high numbers (> 10,000 cells ■ l~') were ob- served for a brief period. The mussels at these sites sometimes exhibited a slight red colouration, but it was rarely sufficient to affect the marketability of the product. Observations from the Phytoplankton Monitoring Program (Carver, unpubl. data) suggested that M. rubrum is a common component of the phytoplankton along the Atlantic Coast of Nova Scotia, particularly during the winter/spring period. It seems that the local strain of M. rubrum is unusual in that it blooms at low temperatures rather than in warm surface waters. For example, records of M. rubrum from other areas of the world suggest that this species is typically associated with "red tide" blooms during periods of upwelling in the late summer or autumn (Fenchel 1968, Taylor et al. 1971. Crawford 1989). One early survey suggested that M. ruhrum rarely exceeds 1.000 cells • P ' at water temper- atures <10°C (Taylor et al. 1971 ). More recently, however, there have been several references to winter/spring blooms of M. ru- brum in cold/temperate coastal waters (Table 2). Given the ubiquitous distribution of M. rubrum and its associ- ation with "red tides." it is difficult to account for the scarcity of records on red colouration in the digestive glands of shellfish. For example, phytoplankton surveys in Passamaquoddy Bay, N.B. (Wildish et al. 1990). indicated that M. rubrum was the fifth most abundant phytoplankton species from 1987 to 1989. Yet. despite the long history of shellfish harvesting in this region, there are no documented accounts of red colouration. On the other hand, there TABLE 2. Estimates of the abundance and/or contribution of M. rubrum to the phytoplankton or microzooplankton community during the winter/spring period (water temperatures <10°C). Location Period Abundance/Contribution Source North Sea Baltic Sea FLodevigen Bay. southern Norway Saanich Inlet, British Columbia Damariscotta Estuary. Gulf of Maine May 1981 Mar-Jun 1982 May 1985 Dec 1975-Feb 1976 Dec 1981 -Apr 1982 Isles of Shoals. Gulf of Maine Dec-Jun 1986 Passamaquoddy Bay. New Brunswick May 1988 Ship Harbour. Nova Scotia Apr-May 1992 Maximum; 50.000 cells ■ P ' 2% phytoplankton biomass. 10% primary production Maximum: 17.900 cells l' 43-859^ microzooplankton biomass. 8% phytoplankton biomass; maximum: 12.000 cells • I ' Major component of microzooplankton biomass (Dec-May) Maximum: 1.200 cells • I ' in March Maximum: 10.600 cells ■ 1 ' Maximum: 30.000 cells ■ P' Gieskes & Kraay 1983 Leppanen & 'jruun 1986 Dale & P ,il 1987 T:tkahashi & Hoskins IQ";) Revelante & Gilmartin 1987 Montagnes & Lynn 1989 Wildishet al. 1990 This study 200 Carver et al. are ar -dotal reports from the East Coast of the United States. John durst (pers. comm.. 1996) noted that soft-shell clams (Mya arenaria) harvested in Maine during the spring often possess red- coloured digestive glands. Similar characteristics were also ob- served in surf clams (Spisiila solidissima) harvested off New Jer- sey in August 1994. It seems that the probability of these charac- teristics developing depends on several factors other than the absolute abundance of A/, rubrum. One is the relative importance of A/, rubrum in the diet of the shellfish; i.e.. in Ship Harbour in April-May 1992 M. rubrum not only occurred in high concentra- tions, but it dominated the phytoplankton biomass. Second, given the tendency of M. rubrum to concentrate in deeper, colder water (e.g.. Location 1 at Ship Harbour), shellfish grown in suspended culture may be more likely to encounter this species than those harvested from wild beds in the intertidal zone. Third, on the basis of limited evidence, mussels appear to accumulate phycoerythrin more readily than do scallops grown under the same conditions. Thus, there may be species-specific differences in the rate of graz- ing on M. rubrum and/or variations in the ability to metabolize phycoerythrin. Finally, shellfish may depurate phycoerythrin more rapidly in the summer and autumn than during the spring, when their metabolic rates are lower. This may explain why records of M. rubrum blooms in warm waters have not been associated with red colouration, even when they occurred in the vicinity of large- scale mussel farms (e.g., MacKenzie et al. 1986). Although there are few reports of A/, rubrum negatively affect- ing the marketability of commercial shellfish species, the increas- ing nutrient enrichment of coastal waters may promote the growth of this species (Lindholm 1985. Lindholm 1992. Dale 1988). In South African waters, decaying blooms of A/, rubrum caused a substantial deterioration m water quality that, in turn, negatively affected other organisms (Horstmann 1981). Furthermore, al- though there is no evidence to suggest that M. rubrum is toxic, Romalde et al. ( 1990) noted that potentially toxic bacteria (e.g.. Vibrio spp.) may appear in association with A/, rubrum blooms. It is thus possible that this species may eventually become a nuisance to coastal communities and the shellfish industry. .ACKNOWLEDGMENTS This work was funded by the Department of Fisheries and Oceans. Biological Sciences Branch, Halifax, Nova Scotia. In particular, we thank Ms. Brenda Bradford (DFO Science) for pro- cessing the histological sections and Dr. Lawrence Fritz of the Institute of Marine Biosciences (1MB) for photographing M. ru- brum and demonstrating the use of the Image Analysis System. Ms. Cindy Legiadro (1MB) also provided valuable advice on using the Image Analysis System. Ms. Kaija Lind and Mr. John Stairs of Aquaprime Mussel Ranch Ltd. provided mussels and space on their longlines in Ship Harbour, and Mr. Dale Cook of Corkums Island Mussel Farm near Lunenburg supplied the control mussels. The control scallops were obtained from Mr. Peter Darnell of Indian Point Marine Farms in Mahone Bav. LITERATURE CITED Bricelj. V. M. & R. E. Malouf. 1984. Influence of algal and suspended sediment concentrations on the feeding physiology of the hard clam Mercenaria mercenaria. Mar. Biol. 84:155-165. Carver. C. E.. S. Hancock. G. G. Sims & W. Watson-Wright. 1992. Phytoplankton monitoring in Nova Scotia (abstract) In: Proceedings of the Third Canadian Workshop on Harmful Marine Algae. Can. Tech. Rep. Fish. Aqual. Sci. No. 1893, p. 30. Clemens. W. A. 1935. Red 'water-bloom' in British Columbia Waters. Nauire 152:473. Crawford, D. W. 1989. Mesodiniiim rulniim: (he phyloplankter that wasn't. Mar. Ecol. Prog. Ser. 58:161-174. Crawford, D. W. & D. A. Purdie. 1992. Evidence for avoidance of flush- ing from an estuary by a planktonic, phototrophic ciliate. Mar. Ecol Prog. Ser. 79:259-265. Dale, T. 1988. Oil pollution and plankton dynamics. VI. Controlled eco- system experiments in LindaspoUenc. Norway. June 81: Effects on planktonic ciliates following nutrient addition to natural and oil- polluted enclosed water columns. Sarsia 73:179-191. Dale, T. & E. Dahl. 1987. Mass occurrence of planktonic oligotrichous ciliates in a bay in southern Norway. J. Plankton Res. 9:871-879. Fenchel, T. 1968. On "red water" in the Isefjord (inner Danish waters) caused by the ciliate Mesodinium rubrum. Ophelia 5:245-253. Fonds, M. & D. Eisma. 1967. Upwelling water as a possible cause of red plankton bloom along the Dutch coast. Neth. J . Sea Res. 3:458—463. Gieskes, W. W. C. & G. W. Kraay. 1983. Dominance of Cryptophyceae during the spring phytoplankton bloom in the central North Sea de- tected by Hl'l C analysis of pigments. Mar. Biol. 75:179-185. Hewes, C. D. & t) Holm-Hansen. 1983. A method for recovering nan- oplankton from filter-, for idenliflcation with the microscope: the filter- liansfer-freeze (FTP) technique. Linmol. Oceanogr. 28:389-394. Hibberd, D. J. 1977. Observations on the ultrastructure of the crypto- monad endosymbiont of the red-water ciliate Mesodinium rubrum. J . Mar. Biol. Assoc. U.K. 57:45-61. Horstmann, D. A. 1981 . Reported red-water outbreaks and their effects on the fauna of the west and south coasts of South Africa, 59-80. Fish. Bull. Sea Fisheries Insl. Rep. S. Afr. 15:71-88. Hummel, H. 1985. Food intake o{ Macoma baltica (Mollusca) in relation to seasonal changes in its potential food on a tidal flat in the Dutch Wadden Sea. Neth. J. Sea Res. 19:52-76, Kat, M. 1984. "Red" oysters (Ostrea edulis L.) caused by Mesodinium rubrum in Lake Grevelingen. Aquaculture 38:375-377. Leppanen. J. M. & J. E. Bruun. 1986. The role of pelagic ciliates includ- ing the autotrophic Mesodinium rubrum during the spring hloom of 1982 in the open northern Baltic Sea proper. Ophelia 4(suppl):147- 149. Lindholm. T. 1985. Mesodinium rubrum — a unique photosynthetic ciliate. Adv. .\quat. Microbiol. 3:1^8. Lindholm. T. 1992. Mesodinium rubrum — a photosynthetic ciliate. pp. 501-514, In: W. Reisser (ed.). Algae and Symbioses; Plants. Animals, Fungi, Viruses, Interactions Explored. Biopress Ltd.. Bristol, En- gland. Lindholm, T. & A. Mork. 1990. Depth maxima oi Mesodinium rubrum (Lohmann) Hamburger & Buddenbrock — examples from a stratified Baltic Sea inlet. Sarsia 75:53-64. MacKenzie, A. L., H. F, Kaspar & P. A. Gillespie. 1986. Some obser- vations on phytoplankton species composition, biomass. and produc- tivity in Kenepuru Sound. New Zealand, 82-83. N. Z. J. Mar. Fresh- »at. Res. 20:397-+05. McAlice, B. J, 1968. An occurrence of ciliate red water in the Gulf of Maine. J. Fish. Res. Bd. Can. 25:1749-1751. Montagnes, D. J. S. & D. H. Lynn. 1989. The annual cycle of Mesodi- nium rubrum in the waters surrounding the Isles of Shoals, Gulf of Maine. J. Planlilon Res. 11:3-201, Revelanle, N. & M. Gilmartin. 1987. Seasonal cycle of the ciliated pro- tozoan and micrometazoan biomass in a Gulf of Maine estuary. Est. Coast. Shelf Sci. 25:581-598. Robinson, W. E., R. W. Langton & C. C. Boggs. 1989. Chlorophyllous pigment and lipid stores in the digestive glands of inshore and offshore Red-Coloured Digestive Glands in Mussels 201 populations of (he deep-sea scallop Placopecien mcif^tilanuus. Mar. Takahashi, M. & K. D. Hoskins. 1978. Winter condition of marine phy- Ecol. Prog. Ser. 52:181-191. toplankton population in Saanich Inlet, B.C. Canada. 11. Micro- Romalde, J. L., A. E. Toranzo & J. L. Barja. 1990. Changes in baclenal zooplanklon. J. E.xp. Mar. Biol. Ecol. 32:27-37. populations dunng red tides caused by Mesodmium ruhrum and Gym- -y^^^^^^ p j r ^ j Blackhoum & J Blackboum. 1971 . The red-water nodimum calenalum in North West Coast of Spain. J. Appl. BacL ^,,,,3,^ Mesodm.um ruhrum and its -incomplete symbionts'-: a review including new ultrastructural observations. J. Fish. Res. Bd. Can. Satoh. H. & K, Watanabe. 1991. A red water-bloom caused by the au- ""g 391-407 totrophic ciliate. Mesodinium ruhrum. in the austral summer in the fast ice area near Syowa Station. Antartica. with a note on their photosyn- '^'^'^^- ^- "^ ■ ^ '^ Sheath & J. A. Hellebust. 1977. A red tide caused thetic rate J Tok\o Univ Fish 7811-17 ''^ ^^^ marine ciliate Mcsoduuum ruhrum in Passamaquoddy Bay. Smith, B. C. & G. H. Wikfors. 1992. Phytoplankton pigments accumu- including pigment and ultrastruclure studies of the endosymbiont. J. lated by the Arctic Surfclam, Mactromeris polynyma (Stimpson, '''-''' '^" ^'^ '^"" -^4:413-416 1860). J. Shellfish Res. 11:479-483. Wildish, D., J. L, Martin, A. J. Wilson & M, Ringuette. 1990. Environ- Smilh, W. O. & R. T. Barber. 1979. A carbon budget for the autotrophic mental monitonng of the Bay of Fundy Salmonid Manculture Industry ciliate Mesodinium ruhrum J. Phxcol. 15:27-33. during 88-89. Can Tech. Rep. Fish. Aquat. Sci. No. 1760. 123 pp. Journal of Shetlthh Resecirch. Vol 15, No. 2. 203-230, 1996. NEOPLASIA AND BIOTOXINS IN BIVALVES: IS THERE A CONNECTION? JAN H. LANDSBERG Florida Marine Research Institute Florida Department of Environmental Protection 100 Eighth Avenue S.E. St. Petershiirs>. Florida 33701-5095 ABSTRACT In the pa^t 25 years, there has been an increase in the frequency of two major types of cancer in bivalves: disseminated neoplasia and germinonias, which cause debilitation and mortality in shellfish stocks. Disseminated neoplasia is common in softshell clams. Mya arenariu: the cockle, Cerastoderma ediile^ and blue mussels, Mytihis trossulus: and less common in edible oysters, Oslrea eJiitis: macomas, Macoiiw halthica: blue mussels, Mytiliis ediilis. and Olympia oysters, Oslrea conchaphiUi . Gemiinomas occur more frequently in northern quahogs, Merceiuiria mercenaria, and softshell clams, Mya arenaria. Certain geographical locations, especially along the northwest Pacific and northeast Atlantic Coasts of North America and the Atlantic Coast of Europe, are "hot spots" for neoplasia. A genetic susceptibility of bivalves to tumor formation has been suggested, and the etiologies proposed include chemical carcinogens, viruses, and other transmissible agents. However, no clear cause-and-effect relationship has yet been conclusively demonstrated, nor has the potential role of biotoxins as etiological agents been examined. In the past 25 years, there has also been an increase in the frequency with which humans have been poisoned by consuming toxic bivalves. Filter-feeding bivalves accumulate biotoxins produced by toxic microalgal blooms. This study traces the worldwide distribution of paralytic shellfish poisoning (PSPl, diarrheic shellfish poisoning, neurotoxic shellfish poisoning, amnesic shellfish poisoning, and venerupin shellfish poisoning and of the microalgae and bivalve species associated with the poisonings and then compares these distributions with the distribution of neoplasia in bivalves. The incidence of disseminated neoplasia in some affected bivalve species appears to parallel, both spatially and tempo- rally, outbreaks of PSP that are associated with the toxigenic dinoflagellates /l/«a/iJn»m tamarense. A. mimilum. A. fiindxense. and A. cateiu'lla. Shellfish that have accumulated potent saxitoxin and its derivatives (neosaxitoxin and gonyautoxins) produced by these dinotlagellates are highly toxic to humans. The presence of disseminated neoplasia parallels the presence of certain toxin derivatives in both the bivalve and the Atexandrium spp. to which the bivalves are exposed. Disseminated neoplasia is common in softshell clams, M. arenaria. that have apparently been exposed to and have accumulated gonyautoxins, (GTX), and in particular GTXl and GTX4, that are produced by A. tamarense oj A . fundyense . M. mercenaria is apparently not affected by disseminated neoplasia and does not usually accumulate toxins associated with A. tamarense or A . fundyense . Bivalves that accumulate high concentrations of saxitoxin or neosaxitoxin, such as butter clams. Sa.xidomus giganleus: surf clams, Spisula solidissima: sea scallops, Placopeclen magellanicus: and California mussels, Mytilus californiunus. are apparently not affected by disseminated neoplasia or germinomas. In M. arenaria. the incidence of germinomas appears to be related to the distnbution of Atexandrium spp. blooms. In M. mercenaria. however, the distribution of germinomas is not related to those Atexandrium spp. that are commonly associated with PSP. The incidence of disseminated neoplasia and germinomas is not correlated with PSP outbreaks associated with Pyrodmium t>uluimense var. compressum or Gymnudinium cawnalwn. Although the epizootiological evidence presented here for a correlation between dinotlagellate toxin profiles, the deposition of toxins in bivalve tissues, and the presence of neoplasia in such bivalves is circumstantial, it should be investigated in field and laboratory experiments. KEV WORDS: Bivalves, cancer, neoplasia, biotoxins, dinoflagellates, Atexandrium. epizootiology NEOPLASIA AND BIVALVES further discussed. Other types of neoplasia that have been docu- Occurrence and Type mented, and are often confused with disseminated neoplasia, are gill carcinomas in Macoma balthica (L.) (Christensen et al. 1974, Since the late 1960s, two main types of neoplasia in bivalves Farley 1976a) and epithelioma-like conditions in Australian rock from marine and estuarine locations around the world have been oysters, Crassostrea commercialis (Iredale and Roughly) (Wolf reported with increasing frequency. The first type, disseminated 1976), neoplasia, affects some 15 species of bivalves (Tables la, 2a, 3a, The second most common type of bivalve neoplasia, germino- and 4a) and can cause heavy mortalities (Elston et al. 1992). In mas or gonadal tumors, affects 10 species and one hybrid (Table disseminated neoplasia, tumor cells are initially found along with 5a). Tumors result from proliferation of the germinal epithelium, normal hemocytes in the circulating hemolymph. As the disease often completely filling the lumen of both male and female gonads progresses, abnormal cells proliferate throughout the blood sinuses (Hesselman et al. 1988, Peters et al. 1994). In germinomas, the and connective tissue of the visceral mass, muscle, and mantle affected gonadal follicles are filled with abnormal, hypertrophic (Peters 1988). The pathogenesis of disseminated neoplasia is sim- cells. Metastasis to the circulatory system occurs in advanced ilar to that of vertebrate leukemia in the sense that the circulating stages (Elston et al. 1992). tumor cells rapidly divide, ultimately invade the connective tissue. Reports discussed here are based on verification of both kinds and in advanced stages, kill the host (Miosky et al. 1989). In of neoplasia in the Registry of Tumors in Lower Animals (Rll^A, bivalves, neither the ontogenesis of normal hemocytes nor that of Smithsonian Institution, Washington D,C,) according to Peters the neoplastic (presumptively hemocytic) cells is known (Elston et (1988) and Peters et al. (1994) and not necessarily as reported in al. 1992). Disseminated neoplasia in bivalves was reviewed by original papers. Rare reports of neoplasia in a particular species, Lauckner (1983), Mix (1986a. 1986b). Peters (1988). and Elston when based on one specimen from many thousands, may need et al. (1992) and, except for certain pertinent facts, will not be further confirmation (Elston et al. 1992). Consequently, the report 203 204 Landsberg TABLE la. Distribution and prevalence of disseminated neopla.sia in oysters witli suspected etiology or conditions. TABLE lb. Corresponding records of dinoflagellate blooms or shellflsh toxicity events by nearest location and date. Bivalve Species Prevalence ( % ) Toxicity/ and Locality (N = ) Date Etiology Reference Blooms Site/Dale Bivalve Reference C. iirginica Hams Creek 0.02 (5.000) 1960-1967 Genetic Couch 1969, Farley ?A. monilautm 'Chesapeake Williams and Ingle York and James 0.01-0.075 1964-1973 1969a, Otto and Far- P. minimum Bay 1972. Steidinger River (51,000) ley 1976. Fnerman 1970.' 1971. 1993. Seliger et Chesapeake Bay 1976, Fnerman and Andrews 1976 1973 Chesa- peake Bay al. 1975 23 sites Chesa- 0,1 (20,000) 1974-1977 Harshbarger et al peake Bay 1979 Apalachicola 0.27 (373) 1978' Couch and Winstead M momlantm 1978 FL, MS Perry et al 1979 Bay. FL (1979) Pensacola Bay. 0 04 (4.486) 8/78-8/80 Couch (1985) A , monilatiim 1978 FL, MS Perry et al 1979 FL fish kills Mobile Bay, AL 0.13 (2.336) Pascagoula Har- 0.44(2,461) bor, MS 0. conchaphila Yaquina Bay, 7.0 C) 1961-1970 Farley and Sparks 1970. PSPA. ca- 1973 Yaquina M galloprovin- Nishitani and Chew OR 0.0-2.0(2,349) ■M975 Mix 1975a. Mix 1975b, Mix et al. 1977 lenella Bay, OR cialts 1988. Taylor 1984 T. chilensis Chiloe, Chile 2,0(4) 2/78 Pristine Mix and Breese 1980 DSPD acuw 1980, 1984. A. aier. M Lembeye et al. 1993 1.0(100) waters PSP? 1991 Reloncavi estu- ary. Jacaf fiord. Chonos archipelago chilensis New Zealand 1 case? 7 Peters 1988 0. edulis Mali-Ston. 20.0-30.0 C) 1975 Heavy Alderman et al. 1977 L- polyedrum. 1980-1^, 1989 Marasovic et al. Croatia mortality A mimttum 1995 Galicia. Spain up to 35.0(7) 1975 PSP A minulum 1976 Galicia M. edulis Luthy 1979 Brittany, France 0.50 (69.476) 1975-1981 C. gigas Balouet et al. 1986 P. minimum 1986 M edulis. Berland and Grze- was - ve PSP A . minulum 1988-1990 0 edulis byk 1991. Erard (4.500) DSP D. acula. D. acumi- nata. D. sac- cuius 1983-1990 Brittany Le-Denn 1991, Belin 1993 of a germinoma in Mytilus trossulus Gould. 1850 in British Co- lumbia (Cosson-Mannevy et al, 1984) and reports of a dissemi- nated neoplasia in Crassostrea rhizophorae in Brazil (Nasimento et al. 1986) and in Crassostrea gigas (Thunberg) in Japan (Farley 1969a) have not been included. Neoplasia Distribution in Bivalves The distribution and prevalence of bivalve neoplasia by type and by species are shown in Tables la to 5a. Neoplasia is common mostly in temperate regions (Figs, 1 to 3). particularly in north- eastern and northwestern North America, the European Atlantic. the North Sea, and Scandinavia, A few cases have been docu- mented in the Gulf of Mexico, Reports of neoplasia are rare in Australasia and the Mediterranean except for the Adriatic Sea near Croatia, In South America, only one case has been documented (.Mix and Breese 1980). In the Middle East, Central America, Africa, and Asia, no reports are known except for one uncon- firmed case in Japan (Farley 1969a), Differences in the predisposition of bivalves to neoplasia are apparent in some families, genera, and species (Tables la to 5a). Oysters, mussels, clams, cockles, and macomas are affected, whereas scallops are not (or are rarely) affected. Oysters are heav- ily affected iOstrea edulis L, and Osirea conchaphila (Carpenter, 1857]), lightly affected {Crassostrea virginua and Tioslrea chi- lensis), or unaffected (C, gigas) by disseminated neoplasia (Table la). Both disseminated neoplasia (Table la) and gcrminomas (Ta- ble 5a) have been found in C. virgiiiica in the Chesapeake Bay but were more common in the 1960s and 1970s than recently. Among the clams, Saxidomus giganteus (Deshayes) and Spisula solidis- sima (Say) are apparently unaffected by either disseminated neo- plasia or germinomas. The northern quahog. Merceiiaria merce- iiaria (L.). is unaffected by disseminated neoplasia (Table 2a) but is affected by germinomas (Table 5a). Mya arenaria L. is heavily affected by both types of neoplasia (Tables 2a and 5a), Blue mus- sels are affected by disseminated neoplasia in some geographical regions but not in others. Along the Pacific Coast of North Amer- ica, M. trossulus is heavily affected and Mytilus californianus (Conrad) is unaffected by disseminated neoplasia (Table 3a). There have been no reports of disseminated neoplasia in Mytilus edulis L. from the northeast Atlantic Coast (North America) or in Neoplasia and Biotoxins in Bivalves 205 TABLE 2a. Distribution and prevalence of disseminated neoplasia in clams with suspected etiology or conditions. TABLE 2b. Corresponding records of dinoflagellate blooms or shellfish toxicity events bv nearest location and date. Bivalve Species Prevalence ( ^i ) Toxicity/ and Locality (N = 1 Dale Etiology Reference Blooms Site/Date Bivalve Reference M arenaria Freepiin. Harp- 10 41 (440) 1967-1977 After oil spill in Yevich and PSP A lama- 1972 York Har- M. edulis Twarog and Ya- swell Neck, 1971 Barzcsz 1976. rense bor. ME M. arenaria maguchi 1975 ME Yevich and Barzcsz 1977 Jones Creek 12 0(50) 9/72 No obvious en- Farley 1976a A lamarense 9/10 1972 An- M. arenaria Hartwell 1975, Annisquam vironmental first reported nisquam M edulis Twarog and Ya- River. MA relationship or viral etiol- ogy outbreak of PSP in the region at same time River, Essex, and Eastham MA A. irradians maguchi 1975. Farley 1976a. Anderson et al 1982 Bourne. MA; 0,0-64 0(1.325) 1-9/76 Highest '7c at oil Brown et al. PSP toxin 4-9/76 western M edulis Hurst 1979 Searsport. spill site,' 1976. 1977 closed shell- ME ME; Quonset. Viral fish beds RI (10 sites) Allen Harbor. 20.0^0 0 (3.500) 7/77-3/79 M. mercenaria Cooper et al PSP toxin 1979NarTagan- M edulis Anderson et al . RI and M ballh- tea - ve 1982a, Coo- per et al, 1982b closed shell- fish beds sett Bay, RI 1982 New Bedford 73 2 NBH (407) 1/82-5/83 ?PCBs* Reinisch et al A lamarense 1987-1988 Borkman et al Harbor 34 3 NBH (886) 5/86- 10/87 Environmental 1984. Leaviit P mmimum NBH and 7 1993. Pierce and INBHI. Liltle 17 0 LBB (881) factors et al 1990 other stations Turner 1994 Bultermilk Bay (LBB); Buzzard Bay. MA Long Island 45.0-60 0 (3.963) 6/83-3/84 Unresolved Brousseau 1987. PSP A lama- 1982-1983 M edulis Schrey et al. 1984. Sound (3 64 3(2.121) 10/88-12/89 Brousseau rense Long Island Nuzzi and Waters sites) Milford and Baglivo P. minimum 1986-1989 1993, Wikfors Point. CT 1991a. Brous- seau and Baglivo 1991b. Brousseau and Baglivo 1994 Long Island Sound. NY and Smolowitz 1993 Chesapeake 0.0-65.0(3,584) 12/83-5/85 Was very rare Farley et al P minimum 1978 Sehgeretal 1979, Bay. MD (6 0 0-78.0 (■') 1990-1995 in this loca- 1986. 1992 Harding and sites) tion pnor to 1984 McLaughlin et al 1996 Coats 1988, Mar- shall 1995 Shrewsbury 0.0-19.0(1,200) 9/86-8/87 Barber 1990 A. lamarense 1987 Atlantic Cohn et al. 1988. River. NJ DSP D acumh naia City. NJ 1980. 1983 NJ. NY Freudenthal and Jijina 1988 New Bruns- 3 1-31,3(688) 12/85-1/87 Morrison et al PSP A ftmdx- 1986 New M edulis. M Martin et al 1990 wick. Nova 1993 ense ( = A , Brunswick. arenaria Scotia (22 excavatum i NS sites). Canada DSP P Ima 1990 Atlantic NS M edulis Man et al, 1992 Mya truncata Baffin Is. . Can- 1.61 (8561 71986 oil? Neff etal, 1987 ada Rudilapes de- cussaius Galicia, NW 1.3>360) 2-12/93 Villalba et al DSP Dinophysis 1991-1993 M gallopro- Blanco et al, 1995. Spain 1995 spp A- minulum G. carenalum Galicia vincialis Franco et al 1994. Anderson et al, 1989 * PCBs, polychlorinated biphenyls. Mytilus galloprovincialis from the northwest Pacific Coast (North America), whereas there have been a few reports of this cancer in both species in Europe (Table 3a). Four species of macomas and one species of cockle have been reported with disseminated neo- plasia (Table 4a). Scallops in the genera Patinopecten, Pla- copecien. and Argopecten are apparently unaffected by dissemi- nated neoplasia, and only one case of germinoma has been reported in bay scallops, Argopecten irradians (Lamarck) (Table 5a). 206 Landsberg TABLE 3a. Distribution and prevalence of disseminated neoplasia in mussels with suspected etiology or conditions. TABLE 3b. Corresponding records of dinoflagellate blooms or shelinsh toxicity events by nearest location and date. Bivalve Species Prevalence ( fc 1 Toxicity/ and Locality (N = ) Date Etiology Reference Blooms Site/Date Bivalve Reference M- trossulus Yaquina Bay. 7,0 12.0(100) 9/68 2/69 No virus found. Farley 1969b, PSPA. cal- 1973 Yaquma A/, trossulus Taylor 1984. Ander- OR 0,4-9,8 (2.934) 6/76-4/78 Etiology re- Mix 1983, enella Bay M. califor- son 1984. Chiang Puget Sound. 0,0-40,0 (40) 11/86 mams un- Elslon et al. A. catenella 1988 Puget nianus 1988. Nishitani WA 11,0(660) 3/89-2/90 known PAH 1988a. Moore Sound, WA S- giganteus. and Chew 1988 Vancouver 0,0-29,2 (166) 12/80-6/81 levels not etal, 1991. A. catenelh 1980-1982 Crassostrea Island, EC 0,0-45,0 (278) 1988' significant Cosson-Man- 1985-1987 BC g'gai Departure Bay 10 0-36,0 (■') 10/83-9/84 nevy et al, BC, Canada 1984, Bower 1989, Emmett 1984 M. edulis Plymouth, 1 61 (994) 1976-1978 Potentially Lowe and PSP/t fama- 1968, 1990 M edulis Wyatt and Sab- England carcmogenic Moore 1978, rense E England ondo-Rey 1993 Morecombe 0,0-4,3 (4,000) 11/78-8/79 PAHs in sedi- Green and '^F. mtntmum Bay, N, ments* Alderman Wales and E. 1983 England Denmark* 0,2-0,8 (8,720) 10/83-9/84 Viral etiology and mullifac- torial hypoth- esis Rasmussen et al. 1985, Rasmussen 1986 ?S? A. lama- reuse "!P minimum DSP Dwophwus nonegtca 1987 1982 M. edulis Moeslrup and Hansen 1988. Kimoretal, 1985 Furuskar, 0,5 (205) 9/86 Sunila 1987 A. lamarense. 1984 Kononen et al, 1993 Tvarminne, D acttmi' Fmland nata. D. acuta. P. minimum. G. calenalum M. galloprovin- cialis Humboldt Bay, 0,0 (40) 1988 — Elston et al. A. calenella 1988 M. gallopro- Price etal, 1991. CA 1988b, vincialis Anderson et al. Rias de Galicia, 0,6(170) 1986 Gutierrez and PSP C, catena- 1976, 1981, M. gallopro- 1989. NW Spam Sarasquete lum. A. minu- 1984-1987 vincialis Franco et al. 1994. 1986 lum DSP Dinophysis acuta. D. acuminata. D. saccutus 1978. 1981. 1983. 1987. 1991-1993 M. gallopro- vinctalis Berland and Grze- byk 1991. Blanco et al. 1985, Blanco el al. 1995 * Elston et al, 1992 list this record as occumng in M. trossulus. * PAH, polycyclic aromatic hydrocarbons. The bivalves most commonly affected by disseminated neopla- sia and in which prevalences of more than 20% have been con- sistently recorded are M. arenaha in the northeastern United States and Canada. M. trossulus in the northwestern United States and Cerastodenna edule (L. ) in Ireland and France (Tables 2a. 3a, and 4a). Mortalities associated with disseminated neoplasia have been recorded in O. edulis (Alderman 1974). O. conchaphila (Far- ley and Sparks, 1970), M. arenaria (Cooper et al, 1982a. Farley et al, 1986. Leavitt et al. 1990. Brousseau and Baglivo 1991b). and M. trossulus (Cosson-Mannevy et al. 1984). Disseminated neoplasia caused mortalities of up to 78% in M. arenaria in New England. The disease may be contributing to recent population declines of M. arenaria in New England (Brousseau and Baglivo 1991b) and in the Chesapeake Bay (Brousseau and Baglivo 1991b, McLaughlin et al. 1996), Prevalences of disseminated neoplasia generally change sea- sonally and are at their highest between October and March (Coo- per et al. 1982a, Cosson-Mannevy et al, 1984, Farley et al. 1986. Rasmussen 1986, Brousseau 1987, McLaughlin et al, 1996), with minimum prevalences from April to August (Leavitt et al, 1990). Biphasic prevalences have also been noted: a second peak may occur from May to July (Cooper and Chang 1982, Cooper et al. 1982a. Barber 1990. McLaughlin et al, 1996) or from January to March (Leavitt et al, 1990), Low water temperatures may suppress the progression of neoplasia (Appeldoom and Oprandy 1980) and reduce mortality (Brown et al, 1977), In field studies, some species that were apparently unaffected by disseminated neoplasia have been found in the same location as other species that were heavily affected. For example, in north- eastern North America. M. arenaha are heavily affected by dis- seminated neoplasia, whereas M, mercenaria. M. edulis. and C. virginica are unaffected. The distribution of germinomas currently appears to be re- stricted to the East Coast of North America, the southern coast of Ireland, New Zealand, and Arctic Canada (Table 5a). Mercenaria spp, and M. arenaria are heavily affected by germinomas. Al- Neoplasia and Biotoxins in Bivalves 207 TABLE 4a. Distribution and prevalence of disseminated neoplasia in macomas and cockles with suspected etiology or conditions. TABLE 4b. Corresponding records of dinoflagellate blooms or shiiirish toxicity events by nearest location and date. Bivalve Species Prevalence (% ) Toxicity/ and Locality (N = 1 Date Etiology Reference Blooms Site/Date Bivalve Reference M. cakarea Baffin Island. Oil) (519) ■'1986 oil' Neffetal, 1987 Canada M hallhua* 4 0-15 0 3/82-7,89 No apparent Pekkannen 1993 A- tamarense. 1984 Kononen et al 1 993 Tvamiinne, Fin- (1.7481 correlation D acumi- land with pollution nata, P mint- mum. G. ca- lenatttm Macoinu inqm- 5.0 (?) ■'1975 PAH' Farley 1976a PSPA. ra- 1973 Yaquina M. califor- Nishitani and Chew nata and M. ieitella Bay. OR manus 1988 nasitta Yaquina Bay. OR C. edule Cork Harbor 0.0-72.0 2/83-2/85 Environmental Twomey and DSP Dtttophysts 1984 M edulis Jackson and Silke and coast. S (1.356) factors/infec- Mulcahy acumtnata. 1995 Ireland (18 tious disease 1984. D acuta. D sites) M, editlis were Twomey and noi~iTgica - ve Mulcahy 1988a A minutum 1987 Cork Har- bour Gross 1988 Bnttany, France 46.0 en '1983 Reference site Poder et al DSPD actila. 1983-1990 Bnt- M edulis Belin 1993 4 sites 4 1 (752) and site of Amoco oil spill; both 1983. Poder and Auffret 1986 D- acumi- nata. D. sac- culus tany had neoplasia PSP Ale.xan- drium minu- tum P minimum 1988-1990 Bnt- tany 1976. 1986 M edults. 0. edulis Belin 1993 Berland and Grze- byk 1991 • This reference may not be a disseminated neoplasm but a gill carcinoma. though M. mercenaria are distributed along the Atlantic Coast of North America, those with germinomas are more localized south of Rhode Island and are particularly prevalent along the southeast Atlantic Coast. Germinomas only occur in M . iirenana in Maine (Barber 1996). The prevalence of germinomas was highest during the warm summer months (Hessleman et al. 1988, Eversole and Heffeman 1993). Germinomas are less common (Table 5a) than disseminated neoplasia (Tables la to 4a). In some incidences, both types of neoplasia were reported in the same species of bivalve at the same time and from the same location, for example, in M. arenaria in northeast North America, in C. edule in Ireland, and on rare oc- casions, in Macoma cakarea in northern Canada (Yevich and Barszcz 1976. Cosson-Mannevy et al. 1984. Twomey and Mulcahy 1984, Neff et al. 1987, Peters et al. 1994). In this situ- ation, a common causative agent might be indicated. Etiology The etiology of bivalve neoplasia has been postulated to be related to various causative agents, but no clear cause-and-effect relationship or multifactorial sequence of events has yet been es- tablished. Tentative links between sublethal exposure to various pollutants and the presence of neoplasia have been postulated but not conclusively demonstrated. A systematic survey of shellfish during the NCAA Status and Trends mussel watch showed that the prevalence of neoplasia was not strongly correlated with chemical contamination (Hillman 1993). Smolowitz and Leavitt (1996) found no correlation between the distribution of disseminated neoplasia in M. arenaria and pollution in Boston Harbor and Cape Cod Bay, MA. Hydrocarbon deposition associated with oil spills was tenta- tively linked to disseminated neoplasia in New England (Barry and Yevich 1975. Yevich and Barszcz 1976. 1977. Brown et al. 1977. 1979, Gilfillan et al. 1977. Harshbarger et al. 1979. Walker et al. 1981); Yaquina Bay. Oregon. (Mix et al. 1979, Mix 1988); Brit- tany, France (Auffret and Poder 1986, Poder and Auffret 1986); and northern Canada (Neff et al. 1987). The presence of neoplasia was demonstrated in areas where chemical contaminants were ab- sent (Gilfallan et al. 1977) or were present at low background levels (Brown et al. 1977, Mix 1983, Cosson-Mannevy et al. 1984, Emmett 1984, Twomey and Mulcahy 1988a). Conversely, neoplasia was absent in areas where bivalves were exposed to extremely high concentrations of contaminants (Mix 1988). Studies attempting to link the occurrence of neoplasia with contaminants have suggested a correlation between the high prev- alence of neoplasia and pesticide use (Farley et al. 1991 , Gardner et al. 1991b). An increased prevalence of disseminated neoplasia in M. arenaria was associated with and statistically correlated to elevated chlordane levels in the tissues (Farley el al. 1991). In recent epizootics, germinomas were observed in M. arenaria from Machiasport. Searsport, and Dennysville, ME (Table 5a). Herbi cides and other agrochemicals were widely used in the extensive forestry and blueberry industries in the area. Gardner a al. (1991b) indicated that the estuaries at Dennysville had been con- taminated by herbicides in a 1979 accidental spray overdrift during the aerial application of Tordon 101* to adjacent forests. Herbicide contamination was the only identified common denominator at all three sites where M. arenaria with germinomas were found (Gard- ner et al. 1991b). Other field studies could not correlate the dis- 208 Landsberg ^ Disseminaled neoplasia A Germinomas Gynmodinium ni'""*^'^ I Alcxandrium fiindycnse I;*!*;*] Alexandrium lamarcnse ^ Alcxandrium calenella Alexandnum irowilnhmi Alexandrium minutura AJexandrium tamiyavanichi Pyrodinium bahamcnse var comprcssum Figure 1. The distribution of dinoflagellates associated witli PSP and the distribution of disseminated neoplasia and germinomas in bivalves. tribution of carcinogenic pollutants with the development of ger- minomas (Yevich and Barry 1969, Barry and Yevich 1972, Hes- selman et al. 1988). Bivalves have been exposed to various chemical pollutants in laboratory exposures (Rasmussen et al. 1985, Rasmussen 1986), but no disseminated neoplasia or germinomas have been induced (Elston et al. 1992). Exposure to chemicals has induced numerous lesions (Farley 1977, Rasmussen et al. 1985) and. rarely, other types of neoplasia in bivalves. Thirty days after C. virginica or M. edulis were exposed to particulate suspensions or solid sediments from Black Rock Harbor. Bridgeport. CT. benign tumors were documented in the kidney, gastrointestinal tract, gonad, heart, and neural tissue (Gardner et al. 1991a). Evidence for a viral etiology for disseminated neoplasia has only been demonstrated in M. arenaria (Brown 1980, Oprandy et al. 1981, Oprandy and Chang 1983). Normal M. arenaria that were exposed to water that had passed over infected M. arenaria developed neoplasia, thus suggesting that a transmissible agent was involved (Brown 1980). When virus-like particles from M. arenaria with neoplasia were injected into normal M. arenaria, these clams subsequently developed neoplasia. Virus-like particles were then reisolated from the newly induced neoplasia, conform- ing to Koch's postulates (Oprandy et al. 1981). A virus similar to an R^NA tumor virus was isolated from M. arenaria with neopla- sia, and after the injection ot the purified virus into normal M. arenaria. neoplasia was induced. Because the virus was not iso- lated from any of the nonneoplastic samples, it was reasoned that a virus was the etiological agent of disseminated neoplasia (Oprandy and Chang 1983). The chemical 5-bromodeoxyuridine was used to induce retrovirus expression and replication as well as disseminated neoplasia in M. arenaria. Oprandy and Chang ( 1983) suggested that the clam tumor-inducing retrovirus may be endog- enous in the cells of normal M. arenaria. A retrovirus was also found in the hemocytic cells of M. arenaria with disseminated neoplasia (Cooper and Chang 1982). Virus-like particles have been demonstrated in disseminated neoplasia (Rasmussen 1986), and a viral agent has been suggested as the probable cause of neoplasia in mussels (Elston et al. 1988a). However, ultrastruc- tural examinations of tissues from C. edule (Auffret and Poder 1986), O. edulis (Cahour and Balouet 1984), M. arenaria (Farley 1976b, Cooper and Chang 1982, Medina et al. 1993), and M. trossulus (Mix et al. 1979) with disseminated neoplasia have failed to reveal the presence of virus. Since the earlier studies demon- strating retrovirus in M. arenaria, a viral etiology has not been confirmed despite numerous attempts (Elston et al. 1992). A viral etiology in the development of germinomas is also unconfirmed. Intranuclear inclusions have been reported in ger- minoma cells of A^. arenaria (Harshbarger et al. 1979; Hesselman et al. 1988), but electron microscopy of the same tissue, which is deposited at the RTLA, did not reveal virus (Peters et al. 1994). An infectious etiology has also been postulated. Disseminated neoplasia appears to be transmissible if neoplastic cells are in- jected into disease-free bivalves (Farley 1987, Elston et al. 1988b, Twomey and Mulcahy 1988b). However, in several experiments, controls were also diagnosed with neoplasia (Farley 1987, Elston et al. 1988b). Kent et al. ( 1991 ) attempted to transfer disseminated Neoplasia and Biotoxins in Bivalves 209 9 Disseminaled neoplasia J^ Genninomas Dinophysis acuta Dinophysis fortii vaH Dinophysis norvegica ^ Dinophysis acuminala Dinophysis sacculus Dinophysis caudata Figure 2. The distribution of dinoflagellates associated with DSP and the distribution of disseminated neoplasia and germinomas in bivalves. neoplasia by injecting blood from heavily affected M. irossulus intoM. arenana. O. ediilis. O. conchaphila. and other M. irossu- lus. After 152 days, only the injected M. irossulus were showing signs of disseminated neoplasia. BIOTOXINS AND BIVALVES In coastal areas where toxigenic microalgae occur, bivalves pose a public health risk because they accumulate a variety of biotoxins by filter feeding on phytoplankton. Exposures to toxic microalgae are usually acute, and high levels of toxins in bivalves prone to toxin accumulation can be reached within days or after only a few weeks. Biotoxins in shellfish are transferred to humans (and other predators) through consumption. The most common poisonings of humans from the consumption of shellfish are par- alytic shellfish poisoning (PSP). diarrheic shellfish poisoning (DSP), neurotoxic shellfish poisoning (NSP). and more recently, amnesic shellfish poisoning (ASP). Venerupin shellfish poisoning (VSP) has rarely been documented. Biotoxins causing human shellfish poisonings are usually associated with dinoflagellates or, in the case of ASP. with diatoms. In addition to the public health risk associated with eating toxic bivalves, the bivalves themselves may be affected by toxin expo- sure. The accumulation of biotoxins in bivalves varies between species, with geography, with the toxicity of specific dinoflagel- lates, and in the localization of toxins in bivalve tissues. Bivalve feeding behavior may be one of the principal factors controlling toxin levels. Some bivalves show immediate behavioral responses to avoid the consumption of toxic dinoflagellates (Gainey and Shumway 1988, Shumway 1990). Some species typically burrow into and feed on sediments, whereas others filter plankton from the water. Toxigenic dinoflagellates can produce benthic cysts and/or vegetative planktonic stages, so bivalves may be differentially exposed to toxins because of their feeding modes. Although some studies have evaluated the effects of short-term toxin exposure on bivalve behavioral and physiological responses, other effects of biotoxins on bivalve health are generally unknown. Despite the frequent exposure of bivalves to biotoxins, no apparent associated pathological effects have been reported (Prakash et al. 1971). The detrimental effects of dinoflagellates and their toxins on bivalves have only recently been considered (Shumway 1990, Shumway et al. 1990, Wikfors and Smolowitz 1993. 1995. Smolowitz and Shumway 1996). The tissues that accumulate toxins and their different compo- nents are known to vary both geographically and temporally among bivalve species, but the effects of chronic exposures are unknown. The majority of available information is on dinoflagel lates known to be producers of toxins that are lethal or deleterious to mammals. The existence of biotoxins or toxic components that are potentially lethal or sublethal to molluscs should be consid- ered. Recent evidence has shown that dinoflagellates that are ap- parently not toxic to mammals may be pathogenic to bivalves (Wikfors and Smolowitz 1995). 210 Landsberg Proroccntrum lima Proroccntnim minimum Figure 3. The distribution of Prorocenlrum spp. implicated in shellfish toxicity and the distribution of disseminated neoplasia and germinomas in bivalves. ASP The first outbreak of ASP, in 1987 in Prince Edward Island, Canada, occurred after humans consumed toxic bivalves exposed to a bloom of the diatom Pseudo-nitzschia multiseries (Hasle) (Bates et al. 1989). Although the effects of diatoms and their toxins on bivalves have not been well documented, ASP is not considered here to be involved with the initiation of bivalve neo- plasia. For the remainder of this article, only biotoxins associated with dinoflagellates will be considered. NSP NSP associated with brevetoxins has been documented from the Gulf of Mexico, the eastern United States, and New Zealand and is produced principally by toxins of Gymnodinium breve Davis and Gymnodinium spp. (Steidinger 1993, Chang 1995). Given that the known distribution of NSP is restricted to these areas, bre- vetoxins are not considered to play a role in the etiology of bivalve neoplasia. Although brevetoxin is well known for its role in fish kills (Steidini;.;T 1993), its effects on molluscs are less well doc- umented. PSP Since the 1970s, there has been a steady increase in the distri- bution of PSP worldwide (Hallegraeff 1993). Outbreaks of PSP are now common in temperate regions, particularly in North and South America, Europe, South Africa, Japan, and Australasia, and in equatorial regions in the Far East, Central Americas, northern South America, and India (Fig. 1). Outbreaks of PSP are related to a series of factors, including dinoflagellate distributions, environ- mental conditions favoring high concentrations of cells, popula- tion toxicity, levels, bivalve distributions, and differential toxin uptake and accumulation by bivalves (Shumway 1990, Hallegraeff 1993). Many bivalve species accumulate PSP toxins, and these species pose a high public health risk during particular seasons and at certain geographical locations. Dinoflagellate Distribution Dinoflagellate species associated with the production of para- lytic shellfish toxins (saxitoxin [STX] and its denvatives) are Al- exandrium acatenella (Whedon and Kofoid), Atexandnum ca- tenella (Whedon and Kofoid), Alexandnum fundyense (Balech), Alexandrium lusitanicum (Balech), Alexandrium minutum (Halim), Alexandrium ostenfeldii (Paulsen), Alexandrium tama- rense (Lebour), Alexandrium lamiyavanichi (Balech), Gymnodi- nium catenatum Graham, Pyrodinium bahamense var. compres- sum (Bohm), and possibly, Alexandrium monilatum and Lingulo- dinium (= Gonyaulax) polyedrum (Stein) (Steidinger 1993). Some cyanobacteria are associated with the production of STX (Mah- mood and Carmichael 1986), and bacteria have been implicated in paralytic shellfish toxin production (Kodama et al. 1990). How- Neoplasia and Biotoxins in Bivalves 211 TABLE 5a. Distribution and prevalence of germinomas in bivalves with suspected etiology or conditions. TABLE 5b. Corresponding records of dinoflagellate blooms or shelirish toxicity events by nearest location and date. Bivalve Species Prevalence ( % » and Locality (N = 1 Date Arctica islan- dica Newpon. Rl 1 case (?) 0 A. irradians Massachusetts 1 case (?) 9 M- calcarea Baffin Island. 1 case ?1986 Canada C. edule Cork Harbor, 0.15(1.356) 2/83-2/85 Ireland C virginka Delaware Bay, 2 0 (50) 8/69 DE, Chesa- 1974-1977 peake Bay. 0.01 (20,000) MD; Black 0.23 (420) 1985 Rock Harbor. CT T. chtlensis New Zealand 21 cases 7 Etiology Reference Toxins/ Blooms Site/Date Bivalve Reference M. arenana Searsport, Ma- chiasport, Denn>sville, ME; Wash- ington County. ME M. edulis New York M. mercenaria Narragansett Bay. RI; Indian R. Lagoon. FL; Charleston. SC Mercenaria campechiensis Tampa Bay. FL; Indian R. Lagoon. FL; Charleston. SC M. campechien- sis X M. merce- naria Indian R. La- goon. FL; Charleston. SC 10-0-43,0 (^) 1 case 0.23 (1.300) 2.3-2.7 (539) 3.3-31.5(1.263) 6.5 (708) 42.0 (?) 58.0-75.0 (440) 7.7 (26) 11.8(85) 42.0 (?) 58.0-75.0 (440) 21.6(75) >42.0('') 100 0 (440) oil? 6.0-12.5 (2,125) 1971-1976 6.4 (204), <22.0 (?) 32.0-^0.0(300) 1979 oil spill, virus her- bicides 1993 ? 1987 summer 68 summer 69/70 5/85-6/87 9/87-8/88 9/87-10/88 1988-1992 9/86 9/87-8/88 9/87-10/88 1988-1992 9/87-8/88 9/87-10/88 1988-1992 No relation- ship with water quality Peters et al 1994 Peters et al. 1994 Neff et al, 1987. Peters 1988 Peters el al, 1994 Farley 1976a. Harsh- barger et al 1979. Gardner et al, 1986 Peters el al. 1994 Yevich and Barszez 1976, 1977, Brown et al. 1977. Harshbarger et al. 1979. Gardner etal, 1991b. Barber 1996 Peters el al 1994 Yevich and Barry 1969, Barry and Yevich 1972. Hesselman et al, 1988, Ben et al. 1993, Eversole and Heffeman 1993, Ever- sole and Heffeman 1966 Hesselman et al, 1988, Bert et al, 1993, Eversole and Hef- feman 1993, Eversole and Heffeman 1996 Bert et al 1993, Eversole and Hef- feman 1993, Eversole and Heffeman 1996 DSPDmo- 1984 physis 1987 spp, A mtnulttm P minimum 1978 A- lama- 1986-1989 rense A lama- rense D. acumi- nata A. monila- tum A. monila- tum A. moni la- tum M. edulis Hurst 1979 Hurst 1979 Jackson and Silke 1995, Gross 1988 Seliger et al. 1975, Nuzzi and Waters 1993 PSPA 1993 Chang et al. minutum. 1995 A. tama- rense PSPA, 1972 M. edulis Twarog and tama- York M. arenaria Yamaguchi rense Harbor, ME 1975 NY 1986- 1989 1984 1978 Indian R La- goon. FL 1978 Indian R La- goon, FL Nuzzi and Waters 1993 Maranda and Shimuzu 1987, Noms 1983 Noms 1983 1978 Indian R La- goon, FL Norris 1983 212 Landsberg ever, ic majority of outbreaks worldwide have been attributed to dinotiagellates. The taxonomy of Alexaiuliium has been in flux. Balech ( 1995) synonymized Alexundrium excuvauiin with A. uimarense and syn- onymized Alexandrium ibehcum (Balech) with A. mimitum. Ba- lech ( 1994) named a new species A. tamiyavanichi that had been previously identified as Alexandrium cohorticula in the Far East (Kodama et al. 1988. Ogata et al. 1990. Pholpunthin et al. 1990. Wisessang et al. 1991. Han et al. 1992). In this article, the most updated references have been used (Anderson et al. 1994. Balech 1995). 1 acknowledge that some records that are based on the onginal authors" descnptions may be inaccurate. When the taxonomy has changed, the original designation has been noted wherever possible. Figure 1 shows the distribution of blooms of the more common toxic dinoflagellate species associated with PSP. In the majority of cases. PSP outbreaks are associated with A. tamarense. A.fiindy- ense. A. catenella. A. miiuilum, C. catenalum. and P . bahamense var. compression (Fig. 1). Particular species have distinct distri- bution patterns: P. bahamense var. compressum is tropical and is common in Asia and Central America; G. catenatum is common along the West Coast of North America, the European Atlantic, southeastern Australia. New Zealand, southern South America, and Japan; A. tamarense is common in northwestern and north- eastern North America and in Europe. New Zealand. Argentina, and the Far East; and A. catenella is common from Alaska to north-central California, central and southern Chile, southeastern Australia. New Zealand, and South Africa but is rare in southern California and Central America (Taylor 1984. Balech 1995). Dinonagellate Toxicity No natural toxigenic dinoflagellate population has been found to contain all naturally occurring PSP toxin derivatives, so the toxin profile is considered to be characteristic of the dinoflagellate strain (Cembella et al. 1993). Some of the PSP toxin derivatives are highly toxic (as sodium channel-blocking agents in mammals) and include the carbamate toxins, saxitoxin (STX). neosaxitoxin (NEO), and the gonyautoxins (GTXl . GTX2. GTX3, and GTX4). The decarbamoyl analogues (dcSTX. dcNEO. dcGTXl . dcGTX2, dcGTX3. and dcGTX4) and deoxydecarbamoyi analogues (doSTX. doGTX2, doGTX3) are of intermediate toxicity. The least toxic derivatives are the /V-sulfocarbamoyl toxins Bl (GTX5). B2 (GTX6). CI. C2. C3. and C4 (Sullivan 1988. Oshima 1995). GTX1/GTX4. GTX2/GTX3. C1/C2. and C3/C4 are pairs in an epimeric relationship; GTXl. GTX2. CI, and C3 are the a-epimers, and GTX3. GTX4. C2. and C4 are the 3-epimers. Essentially, these pairs are in equilibrium with each other, but different physicochemical conditions can shift the ratio of the a- and P-forms (Shimuzu 1987). In some assays, the epimer pairs are combined because of inconsistent epimerization and are thus represented as a combined mol%. The toxin profiles of the more common dinoflagellate species associated with PSP are different (Table 6). By species, the indi- vidual toxin components (moWc) are quite varied. In P. baha- mense var. compressum, there is a lack of CI to C4 and GTXl to GTX4; in A. minntum. only GTX is present, with high levels of GTXl and GTX4 in strains from Spain and Australia and only GTX2 and GTX3 in strains from France (Table 6); in G. catena- turn, there are zero to trace levels of GTXl to GTX4; in A. ta- marense, there are trace to low levels of STX, Bl, and B2 and high levels of NEO and GTXl to GTX4; in A. fundyense, there are low levels of GTX 1 . GTX2. and GTX4 and high levels of GTX3; and in A. catenella. there are high levels of NEO. GTX4. Bl. and B2 and low levels of GTXl . GTX2. and GTX3 (Table 6). Toxin profiles for A. monilatum are unknown (Schmidt and Loeblich 1979). Toxin composition in dinoflagellate species and strains can vary with geographical range and can be influenced by environ- mental factors or experimental conditions (Cembella et al. 1988, Anderson et al. 1990. Anderson et al. 1994). Alexandrium strains TABLE 6. Toxin profiles of dinoflagellates associated with PSP. Toxin (mol^l Dinoflagellate species P bahamense A. A A- A- A. A. A. A. C. var. tamarense numttitm mimititm calenella fundyense ostenfeldii tamiyavanichi lusitanicum* catenatum compressum STX 0.0-3,2 lrace-2-8 26 8 U 4-23 0 0.2 0,0-15 6 NEO 0-3-30.1 trace-22-8 13.2 0.1-3.8 10.5-68.0 GTXl 0.9-20,3 5.0-45,2 00 trace-3,9 0,6 1,1-3 8 26,0-41,0 GTX2 0.1-23,0 <3-0-15.7 18.0 0,1 1,5 0.6 0,3-3,9 6,0 trace GTX3 0-3-86.0 <3-a-10.8 80-0 trace-0.9 50.1 0.1 2.2-10.2 12.0 trace GTX4 12.1-80.5 28.3-90.0 0-0 trace-26,2 5.1 36.8-72,8 41.0-53.0 0.8 Bl (GTX5) trace trace-35 , 5 7.2-13.3 0.3-20,0 26,0-69,4 B2 (GTX6) trace-57,3 91 6 0.1-36.0 4,0-8.0 Cl 1.2-3-2 0,6-3,1 1/2 2.7 1/2 7.7 0.1-7,5 1.2-11.1 C2 49.0-69.1 15.9-70.9 -^ -h 0.4-2,2 6.3-52.2 C3 0.5-2.3 1,9-2,9 6.3-31.3 C4 0.7-1.8 0.2-10.3 5,1-15.0 30.5-68.4 dcSTX 0.7-3.0 0.1 0.1^.0 0,0-4,5 dcGTX2 0.1 2/3 0,1-9,2 dcGTXJ trace 0.1 + Location Japan. Kurea Australia, France Australia, USA Denmark Thailand Ponugal Australia. Malaysia of isolaie Spain Korea Japan Japan, Spain Reference Lassus et al. Hallegraeff et Erard-Le- Hallegraeff et Bncelj et al. Hansen et al. Wisessang et al. Mascarenhas et Oshima et al. Oshima et al 1989. Lee et al. al. 1991. Denn 1991 al, 1991. Kim 1990 1992 1991, al, 1995 1987. 1990, 1987, Usup et 1992. Franco et al et al 1993 Oshima et al. Oshima al 1995 Kim et al. 1993 1994 1990 et al, 1993 * Considered to be a synonym of 4, minuium by Franco et al. 1995. Neoplasia and Biotoxins in Bivalves 213 can vary from highly toxic to nontoxic (Anderson 1990). The original isolate of A. tamarense from the River Tamar, Plymouth, England, and other strains from La Jolla. CA. were found to be nontoxic (Schmidt and Loeblich 1979). The toxicity of A. tama- rense strains increases northwards along the northeast Atlantic Coast of North America (Maranda et al. 1985, Cembella et al. 1988) and northwards in Japan (Kim et al. 1993). This toxicity gradient in isolates from the more northerly latitudes is a reflection of the increased proportion of the highly potent carbamate toxins (STX. NEO, and GTXl to GTX4) \nA. tamarense (Anderson et al. 1982, Anderson et al. 1994). The proportion of the less toxic Af-sulfocarbamoyl fractions such as Cl , C2, Bl . and B2 is higher in the more southern areas (Anderson 1990, Anderson et al. 1994). The presence of A. tamarense has been documented in southern New England and Long Island, but PSP outbreaks are rarer in these areas than they are in the more northerly regions of New England and Canada (Anderson et al. 1982). Bricelj el al. (1991) also pointed out that blooms of A. tamarense are typically less dense in the southern region of its geographical range, which may explain the relative lack of shellfish toxicity in the Long Island area. Analyses of the toxin composition and morphology of 28 strains oi A. tamarense and A. fundyense indicate that although the two species are interspersed geographically from New Jersey to the St. Lawrence estuary and Newfoundland, Canada, only A. fundyense occurs in the Gulf of Maine (Anderson et al. 1994). The north-south trend in toxicity in these isolates was not as distinct as that described by Maranda et al. (1985). but this finding can be partially explained by the fact that high-toxicity isolates from northern areas were not tested (Anderson et al. 1994). The toxin profiles that are discussed in outbreaks of PSP typ- ically refer to those of the bloom-forming vegetative stages. How- ever, the cysts of Alexandrmm spp. are known to be more toxic than the vegetative cells. When newly formed, the cysts can be up to 1 ,000 times more toxic than the vegetative cells and are 10 times more toxic even after several months of dormancy (Dale and Yentsch 1978). Benthic bivalves such as M. arenaria could there- fore be exposed to high levels of toxins at all times if sediments are filtered during feeding. Toxicity in Bivalves The distribution of paralytic shellfish toxins in bivalves varies among species and individuals. This variation occurs initially be- cause of differences in dinoflagellate bloom duration, density, and inherent toxicity. The exposure of bivalves to paralytic shellfish toxins can result in increased mucus and pseudofeces production, modification of valve activity, change in filtration rate, impaired burrowing activity, and altered byssus production, cardiac activ- ity, and oxygen consumption (Shumway and Cucci 1987, Gainey and Shumway 1988, Shumway 1990). In the presence oi A. ta- marense. M. mercenaria close the shell valves (Shumway 1990). This response may partly explain the absence or low level of PSP in this species (Table 7). Other species, like M. arenaria. retract the siphon (Shumway and Cucci 1987) or, like C. gigas. reduce pumping rates (Dupuy and Sparks 1967) when exposed to A. ta- marense and A. calenella. respectively. PSP toxicity levels for C. gigas are lower than those of Placopecten magellanicus (Gmelin) and Patinopecien yessoensis (Jay) (Table 8), and levels for M. arenaria are lower than those of A/, ediilis (Tables 7 and 9), which may partly be the result of these behavioral adaptations. Further differences in uptake dynamics and detoxification mechanisms, in anatomical localization, and in physiological breakdown or trans- formation mechanisms determine the persistence of the toxins in the bivalve tissue (Shimuzu and Yoshioka 1981, Maruyama et al. TABLE 7. Selected examples of maximum toxicity levels reported in clams and the associated dinonagellate species involved in the PSP outbreak. Toxicity Bivalve Date and Location (figof STXeq 100 g"') Dinoflagellate Tissues Reference M. mercenaria 1972 Eastham. MA; 0 A . fundyenselA . tamarense Whole body Twarog and Yamaguchi 1975. 1975 Monhegan Island, 1.113 A . fundyenselA . Whole body Shumway pers. ME tamarense comm. M. arenaria 1972 York Harbor, ME; 1,726 A . fundyenselA . tamarense Whole body Twarog and Yamaguchi 1975. 1972 Merrimack River 9,600 A . fundyenselA . Whole body Twarog and Estuary, MA; tamarense Yamaguchi 1975. 1972 Essex, MA 3,500 A . fundyenselA . tamarense Whole body Twarog and Yamaguchi 1975 S. solidissima 1981 Phippsburg, ME; 7,934 A. tamarense Viscera Shumway et al. 1988, 1990 Georges Bank. ME 6.423 ?/l. tamarense Whole body White et al. 1993 S. giganleus 1985 British Columbia, Canada 9.600 A. catenella Whole body Chiang 1988 S. niillalli 1980 Campbell Cove, CA 14,000 A. catenella Whole body Pnce et al. 1991 Merelrix merelrix 1988 Indonesia 1,400 P. bahamense var. compressum Whole body Adnan 1993 Callisla cinone 1989 Mediterranean Coast, Spain 200 G. catenatum Whole body Bravo et al. 1990 Arctica islandica 1985 Jonesport, ME; > 1,895 A. tamarense Whole body Shumway et al. 1988. 1990 Georges Bank, ME 1,218 ?A. tamarense Whole body White et al. 1993 214 Landsberg TABLE 8. Selected examples of maximum toxicity levels reported in oysters, scallops, and cockles and the associated dinoflagellate species involved in the PSP outbreak. Toxicity Bivalve Date and Location (ixgof STXeq 100 g') Dinoflagellate Tissues Reference A. irradians 1972 Eastham, MA 2.040 A . jundyenselA . tamarense Whole body Twarog and Yamaguchi 1975 C. virginica 1972 Easlham, MA; 0 A . jundyenselA . tamarense Whole body Twarog and Yamaguchi 1975; 1988 Gulf of St. 214 A . fundyenselA . Whole body Worms et al. Lawrence, Canada tamarense 1993 C. gigas 1972 British Columbia. Canada; 1.900 A. catenella Whole body Chiang 1988, 1980 Mann County. CA; 5.500 A. calenella 'Whole body Nishitani and Chew 1988, 1986 Okeover Inlet. EC, 9.929 ■?A. calenella Whole body Shumway pers. Canada comm. Crassoslrea iridescens 1989 SE Mexico 811 Pyroduuum bahamense var. compressum Whole body Cortes- Altamirano et al. 1993 0 ediilis 1986 Harpswell. ME; 1 .300 A. tamarense Whole body Shumway et al. 1990, 1988 Brittany. France 282 A. ininulum Whole body Belin 1993 Cerastoderma sp. 1986 0bidos Lagoon, Portugal 1.096 G. calenatiim Whole body Franca and Almeida 1989 P. yessoensis 71984 Japan 220.000 ?/\. tamarense Digestive gland Noguchi et al. 1984 P. magellanicus 1978 Bay of Fundy. 150.000 A. tamarense Digestive gland Jamieson and Canada; excavatum) Chandler 1983, 1990 Georges Bank. ME; 14.775 A. tamarense Whole body White et al. 1993, 1992 Bay of Fundy. 6.180 A . fundyense Digestive gland Waiwood et al. Canada 1995 1983, Beitler and Listen 1990. Bricelj et al. 1990. Bricelj et al. 1991. Cembella et al. 1993. White et al. 1993, Cembella et al. 1994, Shumway et al. 1994, Cembella and Shumway 1995). STX was first isolated from toxic butter clams, S. giganieus (Schantz et al. 1957, Schantz 1960), and it and at least 20 deriv- atives (Oshima 1995) in various combinations and concentrations have been associated with PSP. The total toxicity of shellfish meat is usually represented as the integrated potency of all toxins present in the sample and expressed in micrograms of STXeq (STX equivalents) per 100 g (Sullivan et al. 1985, Anderson et al. 1984). Shellfish-monitoring standards have an acceptable safety level of 80 fig STXeq 100 g^ ' in raw shellfish soft tissues, and toxicities above this level are considered to pose an immediate public health risk (Clem 1975). A range of STX toxicity levels is found in different bivalves (Tables 7-9): P. yessoensis. P. magel- lanicus, and Mytilus spp. become highly toxic (Tables 8 and 9); M. arenaria have intermediate toxicity levels (Table 7); and M. mercenaria and C. virginica tend not to accumulate or have low levels of toxin (Tables 7 and 8). In general, toxicity levels in bivalves exposed to the vanous dinotlagellates can range from high lo low: high when exposed lo A. tamarense and A, catenella. medium when exposed to A. fundyense and G. catenatum. and low when exposed to A. mimilum and P. bahamense var. compressum (Tables 7-9). When exposed to A. catenella. maximum toxicity levels (in micrograms of STXeq per 100 g) in bivalves varied from 9,929 in C. gigas. 14,000 in Saxidomus nuttalli. 30,360 in M. trossulus. and 127,000 in Mytilus chilensis (Tables 7-9). Bivalve species have different toxin profiles, primarily because of the toxin profile and toxigenicity of the dinoflagellate species to which they are exposed (Tables 10 and II) and secondarily be- cause of their inherent and differential abilities to accumulate and to bioconvert. depurate, or otherwise modify the various PSP tox- ins. Bivalves exposed to P. bahamense var. compressum or G. catenatum accumulate very low levels of GTX, whereas some species that are exposed to Alexandriiiin spp. accumulate high GTX levels (Tables 10 and II). Different bivalve species acquire totally different toxin profiles when exposed to the same di- noflagellate species (e.g.. A. tamarense: Table 12). Additionally, individuals of the same bivalve species can have totally different toxin profiles, depending on the particular dinoflagellate species and strain to which they are exposed and the location and season of exposure. For example, M. edidis accumulate >20 mol% of the derivatives NEO. GTXl, GTX2, GTX4, and CI when exposed to A. tamarense: >20% of the derivatives GTXl, GTX2, GTX3, and GTX4 when exposed to A. minutum: and >20 mol% of the deriva- tives STX and GTX2 when exposed to A. fundyense (Table 10). Neoplasia and Biotoxins in Bivalves 215 TABLE 9. Selected examples of maximum toxicity levels reported in mussels and the associated dinoflagellate species involved in the PSP outbreak. Toxicity ((ig of STXeq Bivalve Date and Location 100 g-') Oinotlagellate Tissues Reference M. edulis 1472 York Harbor, ME; 10.092 A . fundyensel A. tamarense Whole body Twarog and Yamaguchi 1975, 1472 Merrimack River Estuary. ME; 7,392 A . fundxense/ A- lanuirense Whole body Twarog and Yamaguchi 1975, 1972 Essex. MA; 7,200 A . fiimlyense/ A. tamarense Whole body Twarog and Yamaguchi 1975, 1980 Argentine Sea. Argentina; 50,000 A. tamarense {=A. excavatum) ■.'Whole body Carreto et al. 1985. 1981 SW Norway; 42,000 A. tamarense ( =A. excavation) Whole body Langelandet al. 1984, 1986 Harpswell. ME; 2,100 A. tamarense Whole body Shumway et al. 1990 1990 Georges Bank. ME; 24,417 ?A. tamarense Whole body White et al. 1993 1988 Brittany. France 401 A. minutum Whole body Belin 1993 Mxfihis plantilaius 1986 S Tasmania. Australia; 8,350 G. catenatum Whole body Hallegraeff et al. 1989, 1986/1987 Adelaide. S Australia 2,700 A. minutum Whole body Hallegraeft et al. 1989 M. Irossulus 1978 Puget Sound. WA; 30.360 A. catenella ?Whole body Nishitani and Chew 1988, 1982 British Columbia, Canada; 30.000 A. catenella Whole body Chiang 1988, 1987 Kodiak. AK >5.000 ?A. catenella ?Whole body Nishitani and Chew 1988 M. t>ulloprovincialis 1976 Vigo. Spain; 6.000 G. catenatum Whole body Liithy 1979. 1984 Galicia. Spain 445 A. minutum Whole body Bianco et al. 1985 M califvrmanus 1980 Mann County. CA 16.000 A. catenella Whole body Pnce et al. 1991 Mylilus sp. 1986 NW Portugal 1.600 G. catenatum Whole body Sousa et al. 1995 Mytiliis chilensis 1992 S Chile 127,200 A. catenella 'Whole body Benavides et al. 1995 Chliiromyliliis 1989 SW Mexico 542 P. bahamense Whole body Cortes-Altamirano et al. palliopunclalus var. compression 1993 Perna viridis . 1988 Indonesia 1.054 P bahamense var. compressiim ■'Whole body Adnan 1993 Penui perna 1989 Venezuela 1 .309 G. catenatum ■'Whole body La Barbera-Sanchez et al. 1993 Modiolus modiolus 1990 Georges Bank. ME 5,016 "!A . tamarense Whole body White et al. 1993 Tissue Deposition of Toxins Bivalve toxin profiles vary by geographic region (Tables 7-9), by season, and in the distribution of toxic components in different tissues (Beitler and Listen 1990, Cembella et al. 1993, Cembella et al. 1994, Shumway et al. 1994). Some of these differences are reflected in the ability of bivalves to convert toxins both from highly toxic carbamates (STX, NEO. GTXl. GTX2. GTX3. GTX4) to mildly toxic decarbamoyl analogues (dcSTX, dcGTXl , dcGTX2, dcGTX3, dcGTX4) and vice versa or in the ability to store less toxic A'-sulfocarbamoyI toxins (Tables 10 and 11). The ability to convert carbamates to decarbamoyl derivatives has been demonstrated in S. solidissiina. Protothaai stainiitea (Conrad), Penmidia vemdosa. and Mactra chinensis (Sullivan et al. 1983. Briceij and Cembella 1995. Oshima 1995. Bricelj et al. 1996). Bivalves may therefore have different toxin profiles from those of the dinotlagellate to which they were exposed, and their toxin profiles can vary as a function of time since exposure (Cembella et al. 1994). Depuration times vary between different species. Most species can naturally eliminate PSP toxins within weeks (Shum- way 1990). Pacific oysters, C. gigas. are able to depurate toxins from their tissues in less than 9 wk (Shumway et al. 1990). How- ever, S. giganteus, P. inagellanicus. and S. solidissima are known to retain high levels of toxins for long periods of time (from months up to 3 -I- y) (Shumway and Cembella 1993, Shumway et al. 1994, Shumway pers. comm.). In 5. giganteus. the siphons are the main sites of toxin accumulation (Beitler and Liston 1990), and toxins are stored as STX. NEO, GTX2, and GTX3 (Kitts et al. 1992). In P. magellanicus and P. yessoensis, the majority of the toxins is concentrated in the digestive gland, with toxicity levels in the gills and gonads typically less than 80 (xg of STXeq 100 g"' (Shumway and Cembella 1993). In 5. solidissima, toxicity levels of more than 20.000 |a.g of STXeq 100 g' ' were recorded in the gills (Shumway et al. 1994). Tissue storage of toxins can vary by season and by concentration (Cembella et al. 1994). Differences in toxin accumulation in individual bivalves exposed to PSP ranged from 40 to 3,21 3 (jLg of STXeq 100 g"' in September (White et al. 1993). The bivalve accumulation of particular toxins and the deposi- tion of these toxins in different tissues have been studied during laboratory-controlled exposures (Lassus et al. 1989. Bricelj and Cembella 1995). Bricelj and Cembella (1995) exposed 5. solidis sima to A. minutum even though S. solidissima would not typically be exposed to this dinoflagellate, which is rare in North America. The toxins of the A. minutum strain to which the bivalves were exposed were exclusively GTX1/GTX4 (96.9 mol%) and GTX2/ GTX3(3.1 mol<70. After 40 days, the deposition of GTXl /GTX4 in the gills and viscera of the bivalves had declined to less than 5.0 216 Landsberg TABLE 10. To\iii concentrations (niol%) in dinoflagellate species and clams and mussels associated with PSP outbreaks. Where possible, concentration ranges have been provided to reflect the dynamics of toxin sequestration, conversion, and depuration. Bivalve/ Carbamates A'-> ulfocarbamoyls Decarbamoyls Dinoflagellate STX NEO GTXl GTX2 GTX3 GTX4 B1/B2 C1/C2 C3/C4 dcSTX dcGTX2/3 Reference M. edulis 1.3-9.7 1.4-50.2 2.0-^5.7 8.8-36.1 2.1-7.0 2.6-26.3 13.5-42.9 Lee et al. A. lamarense 0.3-0.5 1.3-2.2 17.7-20.3 7.0-7.5 1,9-2.1 12.5-13.5 43.6-44.1 1992 M. edulis 0.0-0.5 2.0-9.0 2.0-7.0 8.0-13.0 9.0-51.0 0.0-60 17.0-24.0 Lassus el A. lamarense 0.0 0.3-1.1 1.1-2.1 7.0-23.0 70.0-86.0 2.4^.0 0.8-1.4 al. 1989 M. edulis 38.8^2.4 21.9-30.7 <0.1-30 7 25.9-76.8 Oshima et A. mtnulum 40.5 5.2 1.3 52.9 al. 1990 M. edulis 40.0 10.0 1/4 13.0 2/3 48.0 2.7 Bricelj et A . fundyense 25.8 13.2 0.6 1.5 50.1 5.1 al. 1990 M. Irossulus + + + + + + Shimuzu .M. catenella etal. 1978 M. californianus 60.9 30.4 1-4 8.7 Whitefleet- A. catenella Smith et al. 1985 M. planulatus 0.1-0.3 0.5-3.0 0.3 0.2 3 6-5.2 1.5-2.7 8.9-18.5 55.9-79.4 5.3-16.2 Oshima et G. catenarum 0.2 trace trace 0.8 0.3-0.8 7.5-63.3 36.8-99.7 0.3-1.2 al. 1987 M. galloprovincialis 5.0 00 1/4 0.0 2/3 0.0 38 5-^2.0 42.0 11 0 Anderson et G- caienaium 6.0 20 1/4 2.0 2/3 0.0 .36.0 17.0 al 1989 M. mercenaria 25.4-30.2 12.2-12.6 1/4 9.3-9.6 2/3 46.0-49.6 1.8-3.9 Bricelj el al 1991 A . fundyense 25.8 13.8 8.9 47.8 3.7 M- arenaria + + + + + + + Martin el A . fundyense al. 1990 M arenaria 23.3 16.2 19 2 17.4 15.6 5.4 0.3 2.7 Hurst et al. A. lamarense 1985 R. phillipinanim 0.O-2.0 O.a-20 0.0-5.0 1.0-17.0 2.5-60.0 0.O-5.0 2.0-13.0 Lassus et A. lamarense 0.0 0.1-1.1 1.1-2.1 7.0-23.0 70 0-86.0 2 4-^.0 0.8-1.4 el al 1989 P. viridis 02 1.8 306 13.4 47 49.3 Wisessang A. lamiyavanichi 0.4 0.0 7.0 0.7 8.5 56.4 el al. 1987 P. viridis 46.7 86 25.5 19.1 Oshima et P bahamense var. 15.6 10.5 694 4.5 al 1987 iompressum Spondylus bulleri + + + + + Oshima et P- bahamense var- al 1990 eompressum Spisula sp. 89.0 9.1 4.6 4.1 2.3 0.6 6.8 Hurst et al. A- lamarense 1985 Merelrix casla 0.4 0.1 22.8 17.8 5.9 12.9 27.9 0 3 1.2-1.8 Karunasagar ?A. lamiyavanichi el al 1990 + , present but no value given. mol% toxin concentration, and GTX2 and GTX3 had been con- verted to dcGTX2 and dcGTX3. Exposures of /W. mercenaria to A. lamarense and A. fundyense (Bricelj et al.l991) indicated that M. mercenaria could accumulate toxins (Table 10), even though toxin accumulation may not occur in the field (Table 7). Cells of the high-toxicity A. fundyense isolate were only consumed if sup- plemented with a nontoxic diatom, Thalassiosira weissflogii (Bricelj etal. 1991). Effects of PSP on Bivalves In the short icrm, bivalves are not usually affected by paralytic shellfish toxins (Kao 1993) because their neuromuscular functions operate mainly by voltage-gated calcium channels. STX and its derivatives block only the voltage-gated sodium channels, which function in mammalian nerves and skeletal and cardiac muscle fibers (Kao 1993). High levels of STX are therefore typically not considered to be lethal or pathogenic to bivalves (Prakash et al. 1971 ). However, the effects of the chronic exposure of bivalves to STX and its derivatives are unknown. Paralysed M. arenaria were reported during the PSP outbreak in western Maine and Massa- chusetts in 1972, whereas toxic M. arenaria in eastern Maine and Canada showed no effects (Prakash et al. 1971). Paralysed M. arenaria are seen in Maine regularly (Shumway pers. comm.). Morbidity and mortality of shellfish were associated with a PSP outbreak in eastern England in 1968 (Adams et al. 1968. Ingham et al. 1968). Eighty percent of C. gii>as that had been exposed to 10 X 10" A. momlalum cells 1 " ' died within 48 h (Sievers 1969). The various combinations of the individual toxins described above determine the toxic potential of the shellfish to humans and the physiological damage expected to occur in the bivalves in which they accumulate. The total STX toxicity is the most impor- tant measure for public health concerns, yet the relative propor- tions of these toxic derivatives and their distribution in different tissues are not always considered. The long-term effects of these exposures on molluscan health needs critical evaluation. Neoplasia and Biotoxins in Bivalves 217 TABLE 11. Toxin concentrations (mol%) in dinoflagellate species in scallops and oysters associated with PSP outbreaks. Where possible, concentration ranges have been provided to reflect the dynamics of toxin sequestration, conversion, and depuration. Bivalve/ Dinoflagellate Carbamates ,V sulfocarbamovls Decarbamoyls STX NEO GTXl GTX2 GTXJ GTX4 B1/B2 C1/C2 C3/C4 dcSTX dcGTX2/3 Reference P. magellankiis 20,0 1,0 3,0 58,0 11,0 < 1 ,0 Fix Wichmann A. tamarense 0,0 11,0 9,0 9,0 41 0 ,30 0 etal, 1981 (=A, excavalum) P. inagellanictis 11,5 8.4 4,6 39,2 23.8 2.3 7.7 + Hurst et al. A. tamarense 1985 P. yessoensis 2,7 34.2 5.4 0.6 0.4 76.8 0.3 2.3-4,8 Oshinia et al A. catenella 1990 P. maximus 0.0-1,5 0.0-2.0 0,0-5,0 1,5-29 0 4.0-40,0 0.5-3.5 3. 0-2 1.0 Lassus et al. A. tamarense 0,0 0.3-1.1 1 1-2.1 7,0-230 70,0-86.0 24-^.0 0.8-1.4 1989 C. gigas 0.0-2.0 2.0-70 00-3 5 2.0-22 U 1,5-46.0 0.0-^.0 7.0-35.0 Lassus et al. A. tamarense 0.0 0.3-1.1 1 1-2 1 7 0-23 0 70.0-86,0 2,4-^.0 0.8-1.4 1989 C. gigas + + + + + + + Onoue et A. catenella al, 1981 C. gigas 0.2 0.0 0.7 0.2 0.1 40 3.1 10.0 79.8 2,1 Oshima et al. G. catenatum trace trace 0,8 0.3-0.8 7.5-63.3 36.8-99.7 0.3-1.2 1987 Crassoslrea cucidlata 0.7 0,0 13,5 52.5 10 1 4,7 14.2 0.0 42 Karunasagar ?A. tamiyavanii'iu et al 1990 DSP DSP is associated with the consumption of shellfish that have been exposed to the dinoflagellates Dinophysis spp. (Fig. 2) and Prorocentritm lima (Ehrenberg) (Fig. 3). DSP outbreaks are most commonly reported in temperate areas in Europe, the Far East, South America, and Australasia (Fig. 2) (Lassus and Marcaillou- Le Baut 1991, Aune and Yndestad 1993). Recently. DSP was documented in eastern Canada (Quilliam et al. 1993). Bivalves currently implicated in DSP outbreaks are M. edulis, Mytilus cor- iisciim. M. galloprovincialis . P. yessoensis. Chlamys nipponensis. Tapes japonica, Gomphira melanaegis. M. mercenaha. Aiila- comya ater. and M. arenaria (Lassus and Marcaillou-Le Baut 199L Lembeyeet al. 1993). Dinophysistoxins (DTXs) (DTX-1. DTX-2. and DTX-3) and okadaic acid (OA) are the major toxins currently known to be involved with DSP. DTXs have been found in Dinophysis ciaimi- nata (Claparede and Lachmann), Dinophysis acuta (Ehrenberg), Dinophysis caudata (Saville-Kent), Dinophysis fortii (Pavillard), Dinophysis norvegica (Claparede and Lachmann I, Dinophysis saccidns (Stein) (Lee et al. 1989), and P. lima (Marr et al. 1992). OA, which is found in some benthic dinoflagellates in tropical regions (Steidinger 1993) and is suspected to have a role in cigu- TABLE 12. Comparative toxin profiles of selected bivalves after exposure to .4. tamarense. Where possible, concentration ranges have been provided to reflect the dynamics of toxin sequestration, conversion, and depuration. STX GTX1/GTX4 GTX2/GTX3 Bivalve (mol%) (mol%) (mol%) C. gigas 0.0-2.0 0.0-7.5 3.5-68.0 M. edulis 0.0-9.7 0.0-72.0 17. 0-64. 0 P. maximus 11,5 0.0-8.5 5.5-69.0 P. magellanicus 11.5-20.0 < 1.0-7.6 0.0-81.8 Spisida sp. 89.0 4.6 6.4 R. phillipinarum 0.0-2.0 0.0-10.0 3.5-77.0 M. arenaria 23.3 24.6 33.0 atera poisoning, has also been found in the planktonic P. lima (Jackson et al. 1993) and Dinophysis spp. (Lassus and Marcaillou- Le Baut 1991). OA and DTX-1 have been experimentally shown to induce skin tumors in mice (Fujiki et al. 1988. Suganuma et al. 1988). The accumulation and metabolism of DTXs in bivalves have not been well investigated, and the effects on molluscan health are unknown. The exposure of mussels to high concentrations of P. lima resulted in reduced filtration rates and was attributed to tox- icity associated with inhibitory or cytotoxic effects (Pillet and Houvenaghel 1995). M. edulis that were experimentally exposed to P. lima accumulated OA and DTX-1 in the hepatopancreas. No mortality was associated with exposure (Pillet et al. 1995). Clear- ance rates of juvenile and adult Argopecten irradians were not inhibited by exposure to toxigenic P. lima, and no mortalities were observed. Toxin saturation levels were attained within the first 2 days of exposure, but toxin retention efficiency was low (Bauder et al. 1996). Dinophysis spp. and P. lima arc widely distributed (Figs. 2 and 3). and the effects of the exposure of bivalves to low-level con- centrations of these dinoflagellates should be investigated. The presence of OA in the planktonic P. minimum has not been con- firmed (see VSP). VSPIProrocentrum minimum VSP has been associated with the consumption of shortnecked clams. Venerupis semidecussata. and Pacific oysters, C. gigas, and was coincidental with blooms of the dinoflagellate Prorocen- iriim minimum m Japan (Akiba and Hattori 1949). VSP is rare, and its true role in shellfish poisonings has been the subject of some discussion. Because of its association with VSP. the widespread distribution of P. minimum (Fig. 3) will be reviewed here. P. minimum is considered to consist of strains that are largely non toxic to humans (Taylor 1984). but toxins that could be pathogenic to bivalves have been isolated (Okaichi and Imatomi 1979). Other shellfish toxicity events associated with P. minimum have been documented in M. edulis in Norway (Tangen 1983), in C. edule and Venerupis decussatus (Silva 1985) in Portugal, and in M. mercenaria in northeastern North America (Freudenthal and Jijina 218 Landsberg 1988). ill Chesapeake Bay. blooms of P . miiunuon appear to be fairly common (Sellner et al. 1993) (Tables lb to 5b) and have recently been associated with shellfish mortalities (Luckenbach et al. 1993). Recent studies have shown pathological effects, inhibition of feeding, and mortality in shellfish exposed to P . minimum (Bar- douil et al. 1993. Luckenbach et al. 1993. Wikfors and Smolowitz 1993, Wikfors and Smolowitz 1995). M. mercenaria and A. irra- dians were fed Proiocentrum micans. P . minimum, and Isochnsis sp. in single-species and mixed-species tests (Wikfors and Smolowitz 1993). M. mercenaria survived well in all experi- ments, but in A. irradians, none of the diets supported good growth. A mixed diet oi Isochiysis and P. minimum caused 100% mortality in 1—4 wk. A. irradians ingested P. minimum, but his- topathological observations showed poorly developed digestive di- verticula, attenuation of the epithelium with abnormal vacuolation and necrosis, and large thrombi in the heart and in the open vas- cular system of the mantle, digestive diverticula, gill, and kidney tissues (Wikfors and Smolowitz 1993). All juvenile oysters. C. virginica, exposed to 100% P. minimum bloom density died within 14 days, and 43% exposed to 33% bloom density died within 22 days, but oysters exposed to 5% bloom density had good shell growth and no mortality (Luckenbach et al. 1993). Wikfors and Smolowitz (1993) suggested that P. minimum produces an enterotoxin that gradually affects absorptive cells, an effect that was indicated by the development of thrombi throughout the vas- cular system. Spat of C. virt^inica exposed to P nunimum had an abnormal accumulation of lipids in the stomach epithelium (Wik- fors and Smolowitz 1995). A Prorocentrum species has recently been implicated in mass mortalities of flat oysters. Ostrea rivularis. in southern China (Yomgjia et al. 1995). The pathology was consistent with a sys- temic toxicosis resulting from the absorption of toxins by the di- gestive gland. Interestingly, the most intense lesion was formed by hemocytes that accumulated in and around the hemolymph chan- nels, infiltrated the walls of the blood sinus, and formed intravas- cular thrombi. This pathology appears to be similar to that found in C. virginica by Wikfors and Smolowitz (1993). These studies suggest that Prorocentrum spp. may induce pathological effects in the hematopoietic system of oysters. UP. minimum produces tox- ins that are important in neoplasia development, could the chronic exposure of oysters to low-level concentrations of P. minimum induce neoplasia of the hematopoietic system? A COMPARISON OF BIVALVE NEOPLASIA AND BIOTOXIN DISTRIBUTION The epizootiology of disseminated neoplasia and germinomas in bivalves appears to closely parallel, both spatially and tempo- rally, the distribution of blooms of dinotlagellate species associ- ated with PSP or VSP (Tables 1-5: Figs. 1 and 3). The correlations noted here are conservative because they reflect only the coinci- dences of acute bloom formations and high concentrations of tox- ins in bivalves They do not take into account the distribution of low levels of dinoflagellate concentrations and thus do not address the potential effect of chronic exposure of bivalves to toxins. These correlations need to be experimentally and statistically ver- ified, A relationship between the distributions of neoplasia and DSP is not currently indicated (Tables 1-5; Fig. 2). My working theory that certain dinoflagellate toxins induce neoplasia in bivalves is based on the currently available toxin profiles of bivalves and dinoflagellates. I recognize that there are gaps in the data and inconsistencies between studies in techniques used; dinoflagellate species, strains, and geographical isolates ex- amined; and time elapsed between bivalve exposures to dinoflagel- late blooms and subsequent analysis of their toxin profiles. How- ever, patterns and trends in the relationship between biotoxins and neoplasia may still be recognized. One of the earliest descriptions of disseminated neoplasia in bivalves in North America mentioned that an outbreak of PSP had been going on in the area at the same time (Farley 1976a) (Table 2). During the first red tide that led to a major PSP outbreak from southern Maine to Cape Ann, MA, in September 1972 (Hartwell 1975. Mulligan 1975). M. edulis and M. arenaria were the most prone to PSP (Tables 7 and 9) and they remained toxic until April 1973 (Hartwell 1975). M. arenaria was heavily affected by dis- seminated neoplasia and germinomas. but M. edulis was refractory (Tables 2a, 3a, and 5a), even in locations where M. arenaria and M. edulis had high toxin levels (Twarog and Yamaguchi 1975) (Tables 7. 9). M. mercenaria and C. virginica did not accumulate toxin (Tables 7 and 8). and they were also refractory to dissemi- nated neoplasia and germinomas (Tables la. 2a, and 5a). PSP outbreaks coincided with several reports of disseminated neoplasia in M. arenaria in Maine during 1972-1975 (Table 2) (Farley 1976a). In August 1986. Morrison et al. (1993) found dissemi- nated neoplasia in 3.1%^ of M. arenaria from Lepreau Harbor, New Brunswick, a month after PSP had been found there in the same species (Martin et al. 1990). Numerous parallel temporal and spatial occurrences of PSP and disseminated neoplasia are shown in Tables 1^. With the exception of the Gulf of Mexico, it appears that the distribution of disseminated neoplasia in bivalves is restricted to comparatively temperate regions in both the northern and the southern hemispheres. Disseminated neoplasia has not been re- ported in Asia. California. Africa, the Middle East, central and northern South America, or the tropics (Tables la to 4a; Fig. I). Thus, for the most part, the distribution of disseminated neoplasia in bivalves more closely parallels the distribution of Alexandrium spp. associated with PSP (Tables 1^; Fig. 1) than that of P. bahamense var. compressum or G. catenatum. PSP outbreaks in Asia are usually associated with P . bahamense var. compressum. and similar associations have recently been reported in Guatemala and Venezuela (Fig. 1). However, toxicity levels in shellfish as- sociated with P. bahamense var. compressum are typically low (Tables 7-9) and are associated with the toxins STX and NEO and their less potent derivatives (Tables 6 and 10). Unlike some Ale.x- andrium spp.. this dinoflagellate lacks toxin derivatives such as GTX that might be potential inducers of neoplasia (Table 6). Cur- rently, there are no documented cases of bivalve neoplasia in areas where P. bahamense var. compressum occurs (Fig. 1). G. calenatum has trace levels of GTX (Tables 6, 10, and 1 1). Shellfish toxicity associated with exposure to G. catenatum typi- cally tends to be low (Tables 7-10) and is usually associated with high levels of the nontoxic components Bl, B2, and CI to C4 (Tables 10 and ID. PSP outbreaks associated with C. catenatum have been reported to occur in Europe, particularly along the At- lantic Coasts of France, Spain, and Portugal, and in Tasmania, Argentina, and California (Fig. 1 ). In some cases, this distribution of G. catenatum parallels that of disseminated neoplasia, but the dinoflagellate has not been reported to occur in northeastern and northwestern North America or in Scandinavia, areas in which there is a high prevalence of disseminated neoplasia. The distri- Neoplasia and Biotoxins in Bivalves 219 butions of neoplasia and G. cateiuitum therefore do not seem to be highly correlated (Fig. 1). In areas where G. calenatum and dis- seminated neoplasia do co-occur, I think that the correlation is more likely to be caused by the presence oi Alexcimiriiini spp.. which co-occurs with C. calenatum in those areas (Fig. 1). Analyses of the toxin compositions of PSP-causing dinoflagel- late species (Tables 6, 10. and 11) show a possible connection between the presence of disseminated neoplasia and exposure to the highly toxic GTX. It is postulated here that the combination of specific toxins will, in some cases, initiate neoplastic development in bivalves. Dinotlagellate species with distributions that parallel that of disseminated neoplasia on a worldwide basis and that have toxin profiles with high levels of GTX are A. lamarcnse. A. ininu- tum, A. calenella, and A. fundyense (Table 6), If there is a relationship between disseminated neoplasia in bivalves and their toxin profiles and concentrations, then high STX or NEO levels do not appear to be as important as other combinations of STX derivatives. It generally appears that when >20 mol'7f of the gonyautoxins GTX1/GTX4 arc present, then disseminated neoplasia is also present (Tables 10-131. When >20 mol% STX or NEO is present, then disseminated neoplasia is generally absent (Table 13). In M. arenaria, after exposure to A. tamarense. >20.0 mol% of STX, GTXI/GTX4, and GTX2/ GTX3 are present (Table 12). However, after exposure to A. fundyense. M. arenaria had low levels of STX and high levels of GTXl, GTX3, and GTX4 (Martin et al. 1990). In this case, the common toxin derivative associated with the presence of dissem- inated neoplasia in M. arenaria appears to be GTX and not STX. Bivalves such as M. lri>ssulus and M. arenaria that store highly potent GTXs are affected by disseminated neoplasia, whereas those species that store STX, such as 5. giganteus, S. soUdissima. P. magellanicus. P. yessoensis, Spondylus butleri. and M. cali- fornianus are unaffected by disseminated neoplasia (Table 13). Recent appearances of Alexandriuni spp. with high levels of GTX such as A. tamarense and A. tamiyavanichi (Balech 1995) (identified as A. cohorticula) in Thailand, Korea, and Japan in the 1980s (Ogata etal. 1990, Pholpunthin et al. 1990, Han et al. 1992) andy4. mimilum in Australasia (Hallegraeff et al. 1991) may fore- shadow the appearance of disseminated neoplasia in predisposed bivalves in these areas. There is a noticeable absence of dissem- TABLE 13. Geographic distribution of neoplasia in various bivalves associated with PSP (high-risk Alexandrium spp.). DSP, and VSP and distribution of toxins (at least > 20 mo\9c). Bivalve PSP DSP Germlnomas Distribution Disseminated Neoplasia Distribution C1/C2 NEO STX GTXI/4 GTX2/3 DTXl OA VSP M. eciulis + NE Nonh America + + Europe + + _ * + + + + -h M. iroisulus 0 NW Nonh America + + + NW North America 9 — ~ + + ND ND ND M. gaUoprovincialis 0 + Europe + - - ? + - + + ND M. californianus 0 0 - + + - - ND ND ND M. arenaria + + + NE Nonh America + + + NE North Amenca ~ — + + + ■? + ? + ND M. truncala 0 + N Canada NO ND ND ND ND ND ND ND C. gigas 0 0 + - - - + ND + + C. virginica + E North America + E North America, Gulf of Mexico ND ND ■' + T. chilensis ? + New Zealand + SW South America. New Zealand ND ND ND ND ND ND ND ND 0. edulis 0 -1- -1- Europe ND ND ND ND ND ND ND ND 0. conchaphila 0 + + NW North America ND ND ND ND ND ND ND P. magellanum 0 0 - - + - + ND + ND P. yessoensis 0 0 + + - - - + -1- ND A irradians + NE North America 0 ND ND + ND ND ND ND ND Mercenana mercenaria + + + E/SE North America. Gulf of Mexico 0 ? + ? + + C edule ? + Western Europe + + + Western Europe ND ND ND ND ND ND + + S. soUdissima 0 0 - - + - - ND ND ND 5. bulleri 0 0 - + + - - ND ND ND A. islandica 4- NE North America 0 ND ND ND ND ND ND ND ND M. ballhicu 0 + + Scandinavia ND ND ND ND ND ND ND ND M. casta 0 ■> + - - + + ND ND ND S. giganieus 0 0 - - + - - ND ND ND + + + , high risk; + + . medium risk; + , low risk; 0. no risk. • STX values for exposures \o A . fundyense are >20.0 mol% (Table 10) (ND, no data). 220 Landsberg inated t'coplasia in bivalves in California, which correlates with and n^ay have been influenced by the absence of A. tamarense in this area, by the presence of low-toxicity G. catenation, or by the fact that California mussels, M. calif ornianiis, retain high STX levels when exposed to A. catcnella (Fig. 1; Table 10). From a public health standpoint, the toxicity of individual. nonconsumable bivalve organs is not usually considered because it is the total toxicity value that is important for safety standards. Toxicities that are reported as micrograms of STXeq per 100 g of shellfish meat (Prakash et al. 1971) are a composite of the total toxicity of the shellfish tissues that are typically consumed by humans. Even though the toxicity of individual organs can be much higher than the overall toxicity of the shellfish meat (Martin et al. 1990). these individual values are only relevant from a human health perspective when particular organs, such as adductor muscles from scallops, are consumed (Shumway and Cembella 1993). From a molluscan health perspective, however, the distri- bution of toxins and derivatives in individual organs may be crit- ical. If neoplastic induction requires a particular period of chronic exposure to one or more toxins, then the deposition of the various toxin derivatives, their concentrations, and their persistence in different organs may play a significant role. At present, both the sites for hematopoiesis and the cellular origin of disseminated neoplasia in bivalves are unknown (Elston et al. 1992). Likely organ sites could include those with open blood sinuses such as the gills, heart, kidney, and brown gland, whereas those such as the adductor muscle and mantle might be less likely. If there is a correlation between the tissue deposition of the highly toxic carbamate gonyautoxins and the prevalence of neo- plasia, then it may be apparent in current bivalve data (Tables 10 and 11). In New England. M. cirenaria is affected by both dis- seminated neoplasia and germinomas (Tables 2a and 5a). Martin et al. (1990) showed that the toxicity of whole M. arenaria extracts had a typical seasonal pattern, with a maximum of 2.103 [x,g of STXeq 100 g ' present in July 1986 in Lepreau Harbor, New Brunswick. Toxicities for some individual tissues were far higher than the total maximum toxicity levels reported (Martin et al. 1990). Levels of approximately 10,000 (xg of STXeq 100 g"' were present in the digestive gland; 6,500 (xg of STXeq 100 g~ ' in the heart, kidney, and brown gland; 500 p.g of STXeq 100 g~ ' in the gills; 300 p.g STXeq 100 g" ' in the gonad; and 120 ^.g of STXeq 100 g " ' in the muscle. Could the deposition of PSP toxins, and particularly the gonyautoxins, in tissues such as the gills, kidney, heart, or brown gland trigger the development of dissem- inated neoplasia? Could the deposition of these same toxins in the gonad trigger germinoma development? After M. cirenaria were exposed to A. fundyense blooms, PSP toxins were transferred rap- idly from the digestive gland to the kidney, where they were retained for extensive periods of time (Martin et al. 1990). Mor- rison et al. (1993) found disseminated neoplasia in M. arenaria from the same area (Lepreau Harbor) as those M. arenaria studied 1 month previously by Martin et al. (1990). Presumably M. are- naria had retained high levels of GTX in susceptible tissues during that period, fhe presence of disseminated neoplasia appears to be more than coincidental, and verification of such a cause-and-effect scenario is critical. In contrast, P. yessoensis are known to be highly contaminated by toxins during PSP outbreaks (Table 8) but are refractory to disseminated neoplasia. Toxin-profile studies of the scallop Pecten maximus show that the accumulation of STX. NEO. and GTX occurs mostly in the digestive gland. The accumulation and sub- sequent transformation of these toxins in the gonad, kidney, and adductor muscle of the scallop P. maximus (L.) lead to an almost complete absence of GTX I and GTX4 in these tissues 15 days after experimental exposure \.o A. tamarense. However, the diges- tive glands still contained GTXl to GTX4 and NEO after 35 days (Lassus et al. 1992). Cembella et al. (1994) reported seasonal variation in toxicity profiles of P. magellanicus tissues. In the digestive glands. GTX2 and CI/C2 were the main components; in the gill. NEO; in the mantle. GTX2 and GTX3; and in the gonads. CI/C2, GTX2. GTX3. and NEO. Levels of GTXl and GTX4 were negligible. The low level or complete absence of GTXl and GTX4 might again explain the absence of disseminated neoplasia and germinomas in scallops (Cembella et al. 1994). The ability to transform toxic PSP carbamates to their corresponding nontoxic decarbamoyi derivatives, as demonstrated by S. solidissima. P. staminea. P. vemdosa. and M. chinensis (Bricelj and Cembella 1995. Oshima 1995, Bricelj et al. 1996), may also correlate with a lack of neoplasia. Again, species with high STX concentrations and low levels of GTX appear to be unaffected by disseminated neoplasia — or at least less affected by disseminated neoplasia than are those species that retain high levels of GTX. Although M. edidis is heavily affected by PSP in northeastern North America, the incidence of disseminated neoplasia has not been recorded in this species in this region. However. M. edidis from northern Europe (England and Scandinavia) are affected by disseminated neoplasia (Table 3a). Several factors could help to explain these geographical differences. In northern Europe. M. cdtdis are more than likely to be exposed to A. tamarense or A. miniiium and. in general, have high GTX levels (>60 mol% GTXI/GTX4) (Table 10). In northeastern North America. M. edii- lis are typically exposed to A. tamarense or A. fundyense. In Maine, where a high prevalence of disseminated neoplasia and germinomas is documented in M. arenaria (Tables 2a and 5a). M. edulis are more than likely to be exposed to A. fundyense (Ander- son et al. 1994). When exposed to A. fundyense. more than 40 moF/f of STX but only about 1 3 molVr of GTX 1 /GTX4 is retained in M. edulis (Table 10). In this situation, the high levels of STX and the low levels of GTX may explain the absence of dissemi- nated neoplasia in M. edulis in this region. I postulate that high levels of GTXI/GTX4 are required to trigger neoplastic develop- ment, li M. edulis are usually exposed lo A. fundyense in Maine and this results in the deposition of low levels of GTX1/GTX4. then the absence of disseminated neoplasia in M. edulis in New England can be explained. The worldwide distribution of germinomas is more localized than that of disseminated neoplasia (Figs. 1-3). If Ale.xandrium spp. are involved in tumor induction, then it might be expected that there would be a parallel distribution of germinomas and disseminated neoplasia in bivalves. In some cases, this situation holds true, as for example, in M. bulthica in northern Canada. M . arenaria in New England, and C. edule in Cork. Ireland. How- ever, in most eases, this situation does not occur (Table 13). The absence of germinomas in bivalves from most of Europe and their rare occurrences along the Pacific Coast of North America suggest that in these areas, toxins from the dinoflagellates A. tamarense. A. fundyense. A. minutum. A. catenella. and G. catenatum are not necessarily involved in germinoma induction. Alternatively, if these dinoflagellates are involved, then the toxin components re- quired for germinoma induction are probably different than those required for the induction of disseminated neoplasia. The high prevalence of germinomas in M. mercenaria and the fact that their Neoplasia and Biotoxins in Bivalves 221 exposure to toxic Alexandrium spp. under natural conditions may not always result in toxin accumulation (Table 7) suggest an al- ternative hypothesis for gcrmmoma induction in this species. There may be a possible correlation between the distribution of neoplasia in bivalves in the Gulf of Mexico and southeastern North America and the distribution of toxigenic A. monilanim or P. minimum. These dinotlagellates have been documented to occur throughout the Gulf of Mexico, the Caribbean, and southeastern and mid-Atlantic North America as far north as the Chesapeake Bay (Steidinger 1993). A potential relationship between the dis- tribution of P . minimum and A. monilatum and neoplasia could also be postulated for the bivalves C . virt^inicu. T. cliilensis. M. menenaria. and Mercenaria campechiensis {Figf<. 1 and 3; Tables 1.5. and 13). The distribution of disseminated neoplasia in oysters appears to be more related to the presence of P. minimum or A. monilanim than to other toxic dinotlagellates in the genus Alex- andrium (Table 1 ). Most accounts of oyster exposure to Alexan- drium spp. report low or no toxicity (Table 8). whereas toxins from Prorocentrum spp. cause pathological effects and have been associated with oyster mortalities (Wikfors and Smolowitz 1993, Wikfors and Smolowitz 1995. Yomgjia et al. 1995). Currently, there is no information on the uptake of toxins from P . minimum ox A. muniUtium by. or their toxicity to. M. mercenaria. although P. minimum cells were found in M. mercenaria in Nassau County. NY in 1985 after an outbreak of human shellfish poisoning (Freudenthal and Jijina 1985). Although there are some mcidences of neoplasia that parallel the distribution of DSP (Fig. 2), there are few records of DSP from the East and West Coasts of North America or the Gulf of Mexico, where neoplasia is prevalent. Dinoflagellate species with associ- ated OA and DTX-1 would appear to be likely candidates for causing tumors in bivalves, yet the existing epizootiology of bi- valve neoplasia does not appear to parallel the known distribution of these dinoflagellates (Fig. 2). However, the focus of this article has been to review the distribution of toxicity outbreaks typically associated with high-density planktonic blooms and acute expo- sure to bivalves. Therefore, if the long-term, low-level exposure of bivalves to OA or DTX is occurring through the continual consumption of Dinophysis spp. and P. lima, then field data com- paring high-density bloom distributions and neoplasia incidence may not be pertment. The influence of anthropogenic chemical carcinogens on the induction of neoplasia in invertebrates has been well investigated (Mix 1986a). Many bivalves have been exposed to highly con- taminated sediments containing chemical carcinogens known or suspected to affect aquatic organisms (Gardner and Yevich 1988). In most cases, a direct cause-and-effect relationship between bi- valve exposure to carcinogens and the induction of disseminated neoplasia or germinomas could not be clearly demonstrated (see Etiology). However, in a few examples, benign tumors developed (Gardner et al. 1991a). One could speculate that if there is a connection between bivalve exposure to particular dinoflagellate toxins and neoplasia, then the neoplasia found in chemical carcin- ogen exposure studies could have been caused by exposure to sedimentary biotoxins. The majority of sediment exposure studies were carried out using sediments from high-risk PSP areas in New England such as Narragansett Bay, RI; Long Island Sound, Black- port, CT; Searsport, Freeport, and Dennysville. ME; and New Bedford Harbor. MA (Yevich and Barscsz 1976. Yevich and Barscsz 1977. Gardner and Yevich 1988) (Table 2b)— all areas where Alexandrium spp. cysts are known to be widespread in the sediments (Anderson et al. 1982, Maranda et al. 1985). It has been documented that the total toxin concentration in cysts of /I. la- mareme is six-fold higher than that in the natural population of vegetative cells (Oshima et al. 1992). Further, these cyst toxins comprised approximately 80 mol9c of GTX compared with ap- proximately 69 mol% of GTX in vegetative cells (Oshima et al. 1992). Theoretically, if these cysts were present in sediments from high-risk PSP areas during the exposure of bivalves to chemical contaminants, then bivalves could also have ingested toxic cysts along with other contaminated sediment particles. There is little or no information about the potential role of natural biotoxins in the induction of tumors in aquatic organisms. OA and DTX-1 produced by Dtnaphysis spp. and P. lima can, in addition to causing DSP. promote tumors in mammals. The fact that bivalves accumulate toxins associated with Dinophysis and Prorocentrum is unequivocal, but the role of OA and DTX in inducing tumors in aquatic animals is currently unknown. It re- mains to be seen as to whether they can trigger neoplasia devel- opment in bivalves. Long-term studies to investigate the relation- ship between toxin exposures and neoplasia should be initiated. A multifactorial etiology of neoplasia development in bivalves could be hypothesized, but before such a step can be made, the role of biotoxins in tumor induction should be defined and clearly demonstrated. In at least one species of bivalve (M. arenaria) known to be affected by disseminated neoplasia, the presence of a retrovirus has been demonstrated (Oprandy et al. 1981). Retrovi- ruses may be endogenous in certain bivalve species and strains such as M. arenaria and Mytilus spp. The proliferation of these viruses could be triggered by exposure to natural carbamate toxins. Carbamate toxins could act directly as mutagens. Different bi- valves may also be predisposed to viral or cellular oncogenes. Genetic differences in species predisposition to neoplasia may be significant (Van Beneden et al. 1993). Genetic susceptibility to germinomas was determined for M. mercenaria. M. campechien- sis. and their hybrids. Hybrids were more affected by germino- mas. which could be explained by decreased genetic fitness (Bert et al. 1993). Although this concept is highly speculative at present, the geo- graphic association of shellfish toxicity events, dinoflagellates, and neoplasia certainly represents strongly circumstantial evi- dence. If gonyautoxins induce disseminated neoplasia, then infor- mation on the chronic deposition of these toxins in different bi- valve species and tissues may be indicative of the differences in species" predisposition to neoplasia. Table 13 shows the geograph- ical distribution of disseminated neoplasia and germinomas. the species affected, and the typically high toxin concentrations (>20 mol%) that some bivalves accumulate. These data can be used to generate a theoretical risk assessment for the geographic distribu- tion of bivalves with disseminated neoplasia (Table 14) and ger- minomas (Table 15). Hypotheses 1 . Disseminated neoplasia and germinomas can be induced in bivalves by toxins produced by dinotlagellates; a bivalve's predisposition to neoplasms is dependent on genetic, behav- ioral, physiological, environmental, and geographic factors that may operate in sequence. 2. Certain species, such as the softshell clam. M. arenaria. and the cockle. C. edule. are affected by both disseminated neoplasia and germinomas, but only in specific geographic locations and at certain times of the year. Other species. 222 Landsberg TABLE 14. Predisposition and theoretical risk of bivalve species to disseminated neoplasia by geographic region and by exposure to dinoflageilate species. PSP p. baha- mense Central/ Asia/ Australia/ .4. A. .4. A. G. var. North South Far New tamar- cate- fundy- minu- cate- compres- VSP Bivalve America America Europe Africa East Zealand ense netla ense tum natum sum p. minimum M. edtilis 0 + + 0 + + 0 0 ' + 0 0 ■'+ + M. planulatits ?+ + ? + + + M. trossutus + + + + + + + + + 0 7 M. galloprovincialis 0 + ? + 0 + 0 ?+ + M. californianus 0 0 0 M. arenarta + + + + + + + + + ?+ + M. trumala + + A- ater 0 0 C. gigas 0 0 0 0 0 0 0 0 0 C. virgtnica + 0 0 ?+ + T. chilensis + + + + ?+ + + + + 0 ?+ + 0- ediilis 0 + + 0 0 + + 0 ?+ + 0. conchaphila + + ? + P. magellantcus 0 0 0 P. yessoensis 0 0 0 TO 0 A. irradians TO 0 0 M. mercenaha 0 0 0 0 C. edule + + + ?+ + + + + + 0 ?+ + S. solidisstma 0 0 0 0 S. bulleh 0 A. islandica + ? + M. balthica 0 + ?+ ?+ + M. casta ?0 0 P. viridis 0 0 S. giganleus 0 0 0 + + + . high nsk, + + . medium risk; + . low risk; 0, no nsk such as the blue mussels. M. edutis and M. trossulus. and the eastern oyster. C . virginica. are rarely affected by both types of neoplasia. Some species, such as M. mercenaha, are apparently affected only by germinomas, and others, such as O. edulis. are affected only by disseminated neo- plasia. The butter clam. 5. giganleus: the Japanese scallop. P. yessoensis: the sea scallop. P. magellanicus: the surf- clam, 5. solidissima: and the California mussel, M. cali- fornianus. are apparently unaffected by either disseminated neoplasia or germinomas. The absence of disseminated neoplasia and germinomas in bivalves from particular geographic regions is likely to be correlated with the absence of dinoflageilate species with high-risk toxins, such as GTXs. or with the resistance to neoplasia of particular bivalve species. Disseminated neoplasia is prevalent in most geographic re- gions where PSP and Alexandrium spp. occur, but only in certain species of bivalves. More specifically, only certain sptc'ies of Alexandrium, such as A. lamarense. A. miniilum, and A. i 14,000 jxg of STX equiv./ 100 g) of PSP toxins in mussels from Puget Sound (J. Wekell, unpublished data, 1977). Since that time. PSP has become a con- stant risk throughout the basin. Predatory snails and other potential human food items besides bivalve shellfish have not been routinely monitored for the presence of PSP or other marine tox- ins. In the Puget Sound basin, PSP levels in shellfish generally begin to rise in May and continue through the summer and early fall months, although high levels of PSP have been noted as late as November and December. Because of their rapacious nature, it is not surprising that PSP levels in moonsnails and other predatory gastropods follow a pattern similar to that observed in other shell- fish in Puget Sound. Butter clams are known to accumulate and retain PSP toxins consistently for long periods along the coasts of Oregon, Washington, British Columbia, and Alaska. Although the levels of PSP in predatory snails found in Puget Sound appear to be low. our September 1994 survey found levels well in excess of the regulatory closure limit of 80 (xg of STX equiv./ 100 g; this indicates that the consumption of these shellfish does represent a potentially serious human health risk. Unlike data reported by Shumway (1995) and White et al. ( 1993a and 1993b) for E. heros ( = Polinices) taken from Georges Bank, we found PSP activity restricted to the viscera of the Puget Sound moonsnail (P. lewissi) in both samplings (fall of 1994 and spring of 1995). In the Georges Bank survey, Euspira PSP levels were considerably higher (>2,500 fxg of STX equiv. /lOO g of tissue) than the toxicities observed in our studies (i.e., <250 (o-g of STX equiv. /lOO g of tissue). In addition, the sampling in this survey was considerably smaller than the sampling conducted in the Georges Bank survey; however, the variation of PSP levels from the September sampling at Agate Passage is similar to that reported by White et al. (1993b) for moonsnails taken from Georges Bank. Nagashima et al. (1995) reported that the trumpet shell [Charonia lampas) collected from the Galician Coast in Spain contained PSP toxins in the digestive gland in a range sim- ilar to those we found in Puget Sound snails. Interestingly, the toxin suite found in the Galician shells was composed of mainly the decarbamoyl derivatives of STX (dcSTX), gonyautoxin2 (dcGTX2), and gonyautoxin3 (dcGTX3), with only minor amounts of GTX2 and GTX3. In order to obtain some understanding about the variation at each collection site, individual shellfish were analyzed. With these constraints, modifications to the AOAC method were necessary because of the nature of the tissues and the amounts available for analysis. Samples of snails from the Double Bluff area were used to test and develop modifications to the PSP assay procedure, i.e., tissue sampling and handling. Moonsnail viscera presented a prob- lem in adjusting the pH. Because viscera consisted of soft tissue, homogenizing with an equal volume of water was unnecessary. For the viscera analysis, 1 part homogenized viscera was mixed with I part 0. IN HCl. but in order to achieve the desired pH 4 with some 1994 samples, considerably more acid was required because a higher than expected initial pH (pH 7.6-8.6) was observed. In order to achieve the desired pH, as required in the AOAC proce- dure, it was necessary to add a more concentrated acid, i.e., 5 N HCl, to the mixtures (10 mL), requiring 20-25 drops or about 1 mL. Because this amount increased the volume nearly 10%, ad- justment to the final calculations was applied. It was noted that the 236 Wekell et al. addition of this acid caused an apparent release of gas, and care had to be taken during the heating step to ensure that material was not lost as the result of frothing and leakage from the tubes. The use of polyethylene centrifuge tubes with fitted screw-cap lids helped reduce these losses and reduce the chances of breakage. With the exception of only one sample, moonsnail viscera from samples collected in April 1995 had normal expected pH values (i.e.. pH 6-7) that did not produce gas when acidified. Neverthe- less, sample 930413-msv-3 had an initial pH of 8. 1 and produced negligible amounts of gas when acidified. It is presumed that this gas is carbon dioxide, because it was both odorless and colorless. Because it is released on acidification, it is probably present as carbonate ion. The high pH encountered suggests that carbonate would be associated with some typical biological alkaline cation, such as calcium, sodium, or magnesium. Accordingly, we expected higher concentrations of these cations in the viscera, which in- cluded the digestive gland/gonadal tissues. Calcium was consid- ered to be the likely cation because of its high concentration in seawater and its involvement in shell synthesis. However, the mineral analyses indicated the opposite of what was expected. That is. the calcium, magnesium, and sodium concentrations on a wet weight basis were approximately one-half of the element con- centrations seen in 1995. in which little or no frothing occurred. Thus, mineral analyses alone are not sufficient to explain the pres- ence of apparent excess carbonate ion. The assessment of health risks posed by the consumption of predatory shellfish and the development of programs for their management will be difficult to implement because our current knowledge, based on this work and other studies in the Puget Sound basin, is limited. To develop an acceptable risk model and program, more information and data concerning the distribution and quantity of these shellfish (seasonally and geographically) will be required. In addition, information about the toxins will be re- quired, for example, levels in specific tissues, total body burden, and seasonal variability. The implementation of a management program will also require detailed information about consumers and their use of these nongame marine invertebrates (NGMI). For example, Carney and Kvitek ( 1991 ) found in their survey that over 50% of the collectors harvesting NGMI were Asian, Korean, or Filipino. The usage and consumption rates of NGMI among these ethnic groups are unknown but are suspected to be much higher than that reported for the whole population (6.5 g/day) (Connie Nakano. Project Coordinator. Asian and Pacific Islander Seafood Consumption Study. Seattle. WA. personal communication. 1995). CONCLUSIONS The results from this survey indicate that Puget Sound preda- tory marine snails accumulate PSP toxins to levels above the reg- ulatory level (80 jig of STX equiv./lOO g). Further, to increase the likelihood of detection, the viscera of the moonsnails appears to be the tissue of choice in analyzing these animals for PSP toxins. Modification of the sample preparation method may be required. Other parts of the moonsnail either do not accumulate the PSP toxins or accumulate it at levels that are below the current AOAC method's detection limit. In addition, these survey data indicate that other predatory and omnivorous snails accumulate PSP toxins. If the recreational or subsistence collection of these species con- tinues or grows, some form of monitoring may be required for the protection of public health. The management of these health risks will require more information on the seasonality, distribution, up- take, and depuration of toxicity in these molluscan species within the Puget Sound basin. Perhaps most important, however, will be more detailed knowledge of the patterns of human collection, preparation, and consumption of these shellfish. LITERATURE CITED Barth. R, H. & R. E. Broshears. 1982. Chapter 9. The mollusks. In The Invertebrate World. Saunders College Publishing, New York. 646 p. Carney, D. & R. G. Kvitek. 1991. Assessment of Non Game Marine Invertebrate Harvest in Washington. Washington Department of Wild- life Report. Department of Wildlife. Washington. D.C. Carriker, M. R. 1961. Comparative functional morphology of boring mechanisms in gastropods. Am. Zool. 1:263-266. Emlen, J. M. 1966. Time, energy, and risk in two species of carnivorous gastropods. Ph.D. Thesis. University of Washington. Seaule. Wash- ington. Hatfield, C. L.. J. C. Wekell. E. J. Gauglitz, Jr. & H. J. Bameu. 1994. Salt clean-up procedure for the determination of domoic acid by HPLC. Nai. Toxins 2:206-211. Keep. J. 1911. West Coast Shells. The Whitaker and Ray Wiggan Com- pany, San Francisco. MacDonald, E. M. 1970. The occurrence of paralytic shellfish poisoning in various species of shore animals along the strait of Juan de Fuca in the state of Washington. M.S. Thesis. University of Washington. Se- attle, Washington. Matter. A, L. 1994. Paralytic Shellfish Poisoning: Toxin Accumulation in the Marine Food Web. with Emphasis on Predatory Snails. EPA 910/ R-94-005, US Environmental Protection Agency. Seattle, Washing- ton, 44 p. Medcof, J. 1972. The St Lawrence rough whelk fishery and its paralyt- ic shellfish problem. Fish. Res. Bd. Can. Man. Rept. Ser. No 1201, 26 p Nagashim.a, Y., O. Arakawa, K. Shiomi & T. Noguchi. 1995. Paralytic Shellfish Toxins in a Trumpet Shell, Charonia lampas. from Spain, p. 74. Abstracts of the Seventh International Conference on Toxic Phy- toplankton, July 12-16, 1995, Sendai, Japan. Prakash, A. J., J. C. Medcof & A. D. Tennant. 1971. Paralytic Shellfish Poisoning in Eastern Canada. Bull. Fish. Res. Bd. Can. Bulletin 177. Fisheries Research Board of Canada, Ottawa, Canada. 79 p. Quayle, D. B. 1969. Paralytic Shellfish Poisoning in British Columbia. Bull. Fish. Res. Bd. Can. Bulletin 168 Fishenes Research Board of Canada. Ottawa, Canada, 67 p. Rice, T 1968 Checklist of the Marine Gastropods from the Puget Sound Region. Issued by OF SEA & SHORE. Port Gamble. Washington. 169 p Rickens, E. & J. Calvin. 1968. Between Pacific Tides. Stanford Univer- sity Press. Stanford, California. Russel, M. E. 1933. West Coast Naticidae. M. S. Thesis. University of Washington. Seattle. Washington. Shumway. S. E. 1995, Phycotoxin-related shellfish poisoning: bivalve mollusks are not the only vectors. Rev. Fish. Sci. 3:1-31. White. A. W.. J. Nassif, S. E. Shumway & D. Whittaker. 1993a. Recent occurrence of Paralytic Shellfish toxins in offshore shellfish in the northeastern United States, pp. 435— 140. In: J. J. Smayda and Y. Shiniizu (eds.) Toxic Phytoplanklon Blooms in the Sea. Elsevier. New York. White, A. W., S. E. Shumway, J. Nassif & D. Whittaker 1993b. Varia- tion in levels of Paralytic Shellfish toxins among individual shellfish, pp. 441^46. In: J. J. Smayda and Y. Shimizu (eds.). Toxic Phyto- planklon Blooms in the Sea. Elsevier. New York. Worms. J.. N. Bouchard. R. Cormier, K. E. Pauley & J. C. Smith. 1993. New occurrences of paralytic shellfish poison toxins in the southern Gulf of St. Lawrence, Canada, pp. 353-358. In: J. J. Smayda and Y. Shimizu (eds.). Toxic Phytoplanklon Bloom in the Sea. Elsevier, New York. Journal of SheUfisb Research. Vol. 15, No. 2. 237-244. 19%. TEMPORAL PATTERNS OF REPRODUCTIVE CONDITION IN THE DOUGHBOY SCALLOP, CHLAMYS (MIMACHLAMYS) ASPERRIMA LAMARCK, IN JERVIS BAY, AUSTRALIA WAYNE A. O'CONNOR'^ AND MICHAEL P. HEASMAN' ^NSW Fisheries Port Stephens Research Centre Salamander Bay, NSW 2316. Australia 'University of Technology . Sydney Department of Applied Biology Gore Hill. NSW 2065. Australia ABSTR.ACT The broodstock reproductive condition of doughboy scallops. Chlamys {Mimachlamys) usperrimu Lamarck, in Jervis Bay. Australia, was monitored fortnightly for 2 y, Gonosomatic index (GSI). gonad weight, and macroscopic gonadal appearance were used to assess changes in the reproductive status of the population. In common with C asperrima in Tasmania, a winter-spnng peak in reproductive activity was observed, although macroscopically npe (mature, ready-to-spawn I individuals were present in most collections. Peaks in reproductive indices occurred in June and August 1992 and in September 1993. during which partial spawning was evident. Male and female development was synchronous, although female scallops maintained higher GSIs and macroscopically higher levels of gonadal development and. at their reproductive peak, had higher calorific values for gonadal tissue. Female gonads were also consistently heavier than those of equivalent-sized males, although adductor muscle weight in males was on average 9% heavier than that of females. Observations of reproductive condition indicate that although the optimal times for the harvest of C. asperrima in Jervis Bay are likely to be similar to that of southern stocks, the presence of reproductively capable individuals at most times of the year in Jervis Bay could be of advantage in the hatchery production of the species and in the provision of embryos for ecotoxilogical studies. KEY WORDS: Chlamys. scallop, reproductive cycle, gonosomatic index, macroscopic index INTRODUCTION The doughboy scallop or fan shell. Chhimvs iMimachkimys) asperruna Lamarck 1819 (Pectinidae), is a gonoehoristic species found subtidally along the southern Australian coast, from Shark Bay, Western Australia, to New South Wales (NSW) (Wells and Bryce 1988: Fig. 1). Capable of growing to over 100 mm in shell height (Zacharin 1994). C. asperrima commonly occurs byssally attached to solid objects in depths of 7-69 m (Young and Martin 1989). C . asperrima has been harvested commercially (Young and Martin 1989) and has aquaculture potential (Cropp 1989, O'Con- nor et al. 1994). and its early ontogenetic stages are used in eco- toxilogical assessments of potential pollutants (Krassoi et al. in press). To understand the life history of a scallop, such as C. asper- rima. manage its fishery, or attempt culture, it is essential to gain an understanding of the reproductive behaviour of the species (Barber and Blake 1991). Recruitment, meat yields, quality of "roe on" product, spawning induction, and the duration of brood- stock availability are all related to reproductive state, which can vary spatially and seasonally. Little is known of the reproductive biology of C. asperrima. and those observations that have been made are largely confined to the southern part of the species dis- tribution (Fig. 1). Rose and Dix (1984; R, Rose pers. comm.) gathered spawnable C. asperrima from the D'Entrecasteaux chan- nel. Tasmania, in October/November. Grant (1971) reported in- dividuals from this area in "almost full roe" in early April, and spawning was thought to occur before July (Zacharin 1986). More recently, Zacharin (1994) monitored reproductive changes in C. asperrima from D'Entrecasteaux Channel and found peaks in re- productive condition in September 1988 and October 1989. In South Australia. Chemoff (1987) caught newly settled spat be- tween February and May. indicating that spawning had com- menced in January and continued until April. To date, no obser- vations of the reproductive condition of populations of C. asper- rima at the northern part of the species range in either NSW or Western Australia have been reported. Visual observations have been the most widely used method of assessing gonadal development in scallops (Barber and Blake 1991). incorporating factors such as relative size, shape, turgor, colour, and acini appearance. Similarly, gonadal indices that ex- press gonad weight as a proportion of total body weight or shell height have been used to define gonadogenic cycles in scallops (Thompson 1977. Bricelj et al. 1987). Both methods were used in this study to show that the reproductive cycle of C . asperrima in Jervis Bay. although less defined, was similar to that of southern populations. MATERIALS AND METHODS C. asperrima were collected from a depth of 15-17 m at a site approximately 500 m northwest of the Murrays Beach boat ramp, Jervis Bay (Fig. 1). This site has a sandy substrate and is well flushed with seawater of salinities in the range of 33-36%r. Water temperatures recorded at the time of scallop collection ranged from 14-23°C. An initial sample of 73 scallops (shell heights, 34-70 mm) was collected by divers in October 1991. Relationships between shell height (central dorsal to central ventral margin) and gonad weight, muscle weight, soft body weight, and gonosomatic index (GSI) were examined by the use of regression analysis. Linear, multi- plicative, exponential, and reciprocal models were fitted to each data set with the "Statgraphics" statistical graphics system (Sta- tistical Graphics Corporation, Rockville, MD), and the model used was chosen on the basis of the highest r" "goodness of fit" value. GSI was calculated by use of the method of Latrouite and Claude (cited in Barber and Blake 1991): GSI = (gonad weight/(soft body weight - gonad weight)) X 100 237 238 O'Connor and Heasman < -( ; Jervis Bay ) { 1 35° 04"/ - ? ■) \ , 30 m \/ X^^Miirrays Beach L.^ P 1 ^ km 1 Depth (m) 150° 44' ( • Collection site ■9s-w^ A ( ' I Western 4 Australia r ' \ South Australia — ( New / ^waies/ Port Stephens "xL f Jervis Bay Vic. X/\ ^ n C. asperrima ^\^>asmania Figure 1. Scallop collection site, Murrays Beach, Jervis Bay, NSW, Australia. Weather permitting. C. asperrima were collected by divers fortnightly from Murrays Beach. Each sample was wrapped in damp jute sacking, packed on ice, and freighted for analysis that evening or the following day. Collections began in October 1991 and continued until October 1993. Shell heights and genders of approximately 30 C. asperrima >50 mm shell height were recorded. The soft body was removed from the shell, and its wet weight was determined to the nearest 0.01 g. The adductor muscle and gonad were then excised and placed in preweighed Petri dishes; any fluid lost after excision was retained withm the dish and included in the weight of the tissue. Where sex was indeterminate or infection with bucephalid para- sites was suspected, samples of gonadal tissue were examined with a stereomicroscope (40x magnification). Bucephalid-infected go- nads were excluded from all analyses reported here. Although sampling was initiated in October 1991, the macro- scopic staging system for C. asperrima was not introduced until February 1992. This delay permitted familiarization with changes in gonadal appearance. Each scallop was assigned to a stage in an arbitrary reproductive index (Table 1), similar to those of Hennick (1970) and Dredge (1981). A numerical ranking was assigned to each macroscopic stage, ranging from 1 for spent scallops to 5 for ripe (mature, ready-to-spawn) scallops. Partially spawned scallops were given a ranking of 4. the same as that of well-developed scallops, because many of these scallops still had relatively full, turgid gonads and, on the basis of observations of induced spawn- ings within hatcheries, were capable of contributing to spawning events. The gross energy value of ripe male and female gonads was determined by bomb calorimetry. Four batches of five gonads from each sex were compared. Scallops chosen for analysis were collected during peak reproductive periods in late August and Sep- tember 1993. Each scallop selected was judged to be ripe on the basis of macroscopic criteria and had a GSl >25%. Each gonad was carefully excised to ensure that little or no tissue from the digestive gland was included. The gonad was then dissected, and the intestinal loop was removed. The remaining gonadal tissue was weighed to the nearest milligram, dried at 105°C for 24 h, weighed again, and stored in a freezer until analysed. Dry matter was determined for each pooled sample (five gonads) before gross energy evaluation with a Parr 1241 adiabatic bomb calorimeter (Parr Instrument Co., Moline, IL). RESULTS All 73 scallops in the initial collection were judged to be ma- ture on the basis of macroscopic observations. Gonadal tissue samples from the smallest of these scallops (34-45 mm) were examined microscopically (lOOx magnification) and found to contain either spermatozoa or developing oocytes. C. asperrima gonad, muscle, and soft body weights all increased exponentially TABLE 1. Macroscopic gonadal staging system for determination of reproductive condition in the doughboy scallop, C, (Mimachlamys) asperrima. Stage Numerical Ranking Score Spent/Immature: Spemi or eggs absent. Gonad flaccid and translucent. Both ascending and descending limbs of intestinal loop clearly visible. 1 Developing 1: Gonads are filling; separate acini are apparent, giving a granular appearance; males and females are distinguishable. 2 Developing 2: Gonad less granular in appearance as acini begin to fill. Gonad increasing in turgor. 3 Developing 3: Very little of the intestinal loop visible, usually only a small portion of the ascending limb at the distal extremity of the gonad. Gonad appears uniform in colour and texture as acini till. 4 Ripe: Gonad uniform in colour, highly turgid, acini not apparent, intestinal loop not visible. 5 Spawning; Gonad uniform in colour; however, flecks or mottling occurs as the result of voided acini. Gonad turgor varies according to the extent of spawning, usually the equivalent of Developing 2 & 3 scallops. 4 Reproductive Patterns in Chlamys asperrima 239 T3 a c o O (•2.402 + 0.056X) r -- 0.90 y-e 2 50 (3 10 y = 30.681 - 0.929 x r^-0.02 a 60 Ci" 40 § 20 35 45 55 65 75 Shell height (mm) 35 45 55 65 75 Shell height (mm) 24 ■o o (-0.859 ■>■ 0.057 X) y 6 10, r2.0. ,94 o 0 , (-2.492 + 0.065 x) = 0.90 35 45 55 65 75 Shell height (mm) 35 45 55 65 75 Shell height (mm) Figure 2. Relationship between .sliell height and gonad weight, muscle weight, soft body weight, and GSI for the initial collection of C (Mi- machlamys) asperrima (October 1991, shell height, 34-70 mm). with shell height, and although no significant relationship was found between scallop size and GSI (Fig. 2). subsequent collec- tions were limited to scallops >50 mm shell height. Larger scal- lops were selected to reduce the possibility of size-dependent vari- ation in GSI (Conor 1972, Grant and Tyler 1983. West 1990) and to improve the efficiency and precision of dissections. Previous studies with hatchery-produced juveniles indicated the onset of sexual maturity at approximately 30 mm shell height (O'Connor et al. 19941. During the following 2 y. 1,612 scallops were examined. Of these. 51.2% were female and 44.47f were male. Genders of 4.3% of scallops collected were indeterminate because of parasitic cas- tration by a bucephalid trematode. A comparison of shell height data for each scallop collection found that variances were highly heterogeneous (Cochrans test, C = 0.056; p < 0.001) and that shell height varied significantly throughout the sampling period (Kruskall Wallace test, x' = 340.2; 5:df = 28; p < 0.001 ) and 24/8/93 (F = 10.7; df = 29: p < 0.01 ), we considered GSI to be largely independent of shell height during this study. In each case, a negative relationship was found, that is, GSI decreased with increasing shell height. Significant variation in GSI was evident between collections (Kruskall Wallace text, x" = 890.3, (// = 51, p < 0.001 ). High mean GSI values (>18'7f ) were generally found from late autumn to early spring (May to September), peakmg in the months of June and August/September 1992 and in September 1993 (Fig. 3). Scallop reproductive condition was considered to be poor, that is, smaller, flaccid gonads with much of the intestinal loop visible (GSI < 18%), from early summer to autumn (De- cember to April) in both years. Macroscopic Staging Changes in mean ranking for macroscopic stages for each col- lection showed a pattern similar to that of GSI and gonad weight (Table 2). Ripe scallops were present in most collections (Fig. 4), with the highest occurrence corresponding closely with peaks in GSI. In addition, ripe scallops were uncommon or absent in the collections of March and April 1992, when the lowest mean GSI 240 O'Connor and Heasman TABLE 2. Correlation matrices for regressed muscle and gonad weights for a 62.5-mm C. iMimachlamys) asperrima (predicted on the basis of regression analysis) for each collection, the mean GSI for each collection, and the mean macroscopic ranking. Parameter Regressed Gonad Weight Mean GSI Mean Macroscopic Gonad Ranking Regressed muscle weight Regressed gonad weight Mean GSI r = -0.21 ip = 0.144, n = 52) r = -0.45 (/> < 0.001. n = 52) )• = 0.91 (p < 0.001, n = 52) r = 0.51 (p < 0.001. n = 43) r = 0.80 (p < 0.001. II = 43) r = 0.84 [p < 0.001. n = 43) figures were recorded. During the study, only \.59c of scallops were found to be spend and no spent scallops were found between August and December in either year. Spawning scallops were also uncommon, comprising only LlVt of the scallops ranked. This may have arisen from difficulties in differentiating scallops in the advanced spawning stages from those in Developing 1 & 2 stages. Despite this difficulty, staging offers a fast, useful alternative to GSI and. importantly, is nondestructive. Macroscopic staging at the time of collection was not observed to increase the incidence of unplanned spawnings in scallops maintained for later experimen- tation. Tissue Weights Examination of variation in tissue weights was not made until each weight had been related to shell height by the use of linear least-squares regression (Daniel and Wood 1980. Dredge 1981). For gonad weights, the equation where y is gonad weight (in grams) and x is shell height (in mil- limeters) was fitted to each fortnightly sample, and the predicted gonad weight for a scallop of 62.3 mm shell height was calculated. The exponential equation was chosen instead of linear or multi- plicative equations on the basis of higher r percentages when fitted to weight data from the initial collection (Fig. 2). A shell height of 62.5 mm was selected because it approximated both the mean and the modal shell height of scallops collected during the study. The same procedure was repeated for muscle and soft body weights. Changes in gonad weight with time of year were found to closely reflect changes found in GSI (Fig. 3). Generally, vanation in muscle weight reflected the inverse of that observed with con- dition indices, although no significant negative correlation was found between regressed gonad and muscle weights (Table 2). Peaks in muscle weight occurred in summer in 1992 and 1993, when the average muscle weighed approximately 5.9 g. nearly double the winter low of 3. 1 g. ojr^r^.,-o>oj-,-oooKO)0>a>moo)ocoo coaDincciu^ojtocDu^intnmcvjcocDojojcDcD ce 0) CL m 20 ._it_ ;; __ :_. 1;:, _. _i : ^, ^_^i^ , , FMAMJJASONDJFMAMJJASO 1992 I 1993 Spent D Developing 2 Spawning n Developing 3 Developing 1 M Ripe Figure 4. Monthly frequency of macroscopic stages of gonadal devel- opment in the scallop C. (Mimachlamys) asperrima from February 1992 to October 1993. Comparison of Results for Males and Females The ratio of males to females in each collection did not vary greatly over the sampling period, and the overall ratio of males to females collected did not differ significantly from 1:1 (x" = 3.78; df = \:p> 0.05). Observations of bucephalid parasitism, which can prevent the macroscopic determination of sex, indicated that it is not exclusive to either sex. Mean shell heights of males and females collected were 62.3 ± 5.1 and 62.5 ± 4.8 mm (mean ± SD). respectively (Table 2). Over the 2-y sampling period, the mean muscle weight of males (4.43 g) was significantly (/; < 0.05) greater than that of females (4.04 g). Conversely, the mean gonad weight in females (2.27 g) was significantly (p < 0.05) greater than that of males ( 1 .97 g; Table 2). However, the average soft body weight of males did not differ significantly ip > 0.05) from that of females (Table 2). Coefficients of variation for shell height or the various tissue weights did not differ greatly between sexes (Table 2), and changes in each measure over the sampling period were synchro- nous (Fig. 5). The macroscopic stage of gonad condition for the sexes also varied synchronously; however, females on average achieved significantly higher rankings (x" = 261.6; df = 5; p < 0.001; Fig. 6). The percentage of females judged to be in ripe or spawning condition over the 2-y period was more than twice that of males. 17.5 and 7.6%. respectively (Fig. 6). A comparison of gross energy in ripe male and female gonads showed males to have significantly (t = 6.9. p < 0.01) less energy. Male gonads averaged 17.9% dry matter and 18.66 MJ Ri PRODUCTIVE Patterns in Chlamys asperrima 241 TABLE 3. Comparison of mean shell height and tissue weights of 716 male and 826 female C. (Mimachlamys) asperrima collected from Jervis Bay between October 1991 and November 1993. Male Female / Parameter Mean ± SD Range cv Mean ± SD Range cv P Shell height (mm) 62.26 ±5.13 31.30 8.2 62.45 ± 4.78 33.00 7.6 0.73 0.46 Soft body weight (g) 13.76 ± 3.68 21.06 26.8 13.45 ± 3.30 22.98 24.5 1.79 0.07 Gonad weight (g) 1.97 ± 0.87 4.98 44.0 2.27 ± 0.98 6.92 43.1 6.46 <0.01 Muscle weight (g) 4.43 ± 1,60 9.65 36 1 4.04 ± 1.43 10.56 35.3 5.04 <0.01 'CV, coefficient of variation (%). kg" ' dry matter, whereas female gonads averaged 19.5% dry matter and 19.65 MJ kg~' dry matter. At the time of collection, the calculated gonad weight for a 62.5-mm female scallop was 5.12 g. and for a 62.5-mm male, it was 4.31 g. On a static basis, this implies that the energy invested in male gonads at that time was 27% less than that in female gonads. DISCUSSION Evidence was found to suggest that an annual reproductive cycle occurs in C. asperrimci from Jervis Bay. Peaks in all repro- ductive indices occurred in winter and early spring, and on the basis of considerable fluctuations in GSl (Dredge 1981: Sause et al. 1987, West 1990, Zacharin 1994), there were indications that some spawning occurred. The scarcity of spent individuals, even during the winter-spring peak in reproductive activity, indicated that either redevelopment is extremely rapid or that partial or drib- ble spawning was responsible for these fluctuations. Evidence for partial spawning can be found in the twin peaks in GSI in 1992, in the protracted settlement period for spat in South Australia (Cher- noff 1987), and in the stepped decline in the gonadal weight of C asperrima from both D'Entrecasteaux Channel in Tasmania (Za- D) O o w 0) 0) o 3 1991 I 1992 I 1993 1991 I 1992 I 1993 CO ■fl) ■D 03 c o O 1991 I 1992 I 1993 1991 I 1992 I 1993 Figure 5. A comparison of monthly mean GSI and regressed tissue weights for male ( — ) and female ( ) C. {Mimachlamys) asperrima (shell height, 62.5 mm). charin 1994) and Jervis Bay. Peaks in reproductive activity in C. asperrima in this study were temporally consistent with reports from Tasmania and with the winter-spring spawning period ob- served for some other tropical and subtropical pectinids in the southern hemisphere (Sause et al. 1987), Grant's (1971) observa- tion of "almost full roed" C. asperrima in April and the October/ November collection of spawnable broodstock by Rose and Dix (1984) coincide with the beginning and end of the main period of reproductive activity in Jervis Bay. Further, the peaks in repro- ductive indices for C. asperrima corresponded closely with those recorded by Zacharin (1994) in the D'Entrecasteaux Channel. Re- calculating GSI in this study to that used by Zacharin (ratio of gonad weight to soft body weight) found that mean GSI varied in a range from 5.9 to 31.8%, which is similar to the range reported by Zacharin (1994; c. 5-33% for males and females combined). Observations of spat occurrence were generally consistent with winter/spring spawnings. With the exception of a small number of juveniles less than 7 mm in shell height collected from Jervis Bay in August 1994 (W. O'Connor pers. obs.), recruitment was not observed by divers in 1992 or 1993 until January or February, when juveniles had grown to 8-10 mm in size. Previous experi- ence with hatchery-produced C. asperrima has shown that spat could grow to 10-mm shell height in 12-15 wk (O'Connor et al. 1994), and thus, spawnings leading to recruitment during this study were thought to have occurred in early spring of the previous year. Despite their similarities, two important differences were evi- 50 1 40 I 30 CD O a> Q. 20 10 I I Female I I Male ^ l=d Spent Spawning D1 D2 D3 Ripe Macroscopic ranking Figure 6. Percent frequency of macroscopic stages for male and fe- male C. [Mimachlamys) asperrima collected from February 1992 to October 1993. 242 O'Connor and Heasman der jetween populations of C. asperrima in Jervis Bay and those in D'Entrecasteaux Channel. First, a distinct difference in maxi- mum size was noted. Tasmanian C. asperrima have been reported to exceed 100 mm in shell height (Zacharin 1986), whereas those collected from Jervis Bay over the past 3.5 y rarely exceeded 80 mm. Second, Zacharin (1994) observed a "resting phase" from January to March in which gonads were completely spent and the majority of individuals could not be sexed macroscopically. Go- nads of C. asperrima in Jervis Bay were smaller and macroscop- ically less developed during this period; however, very few fully spent individuals were found and macroscopically ripe individuals were present in collections from most months. Indeed, C. asper- rima collected during this time have been induced to spawn (W. O'Connor unpubl. data, R. Krassoi pers. comm.). Similar in- traspecific variations in latitudinally differentiated bivalve popu- lations have been reported previously. In general, bivalves from lower latitudes have a smaller maximum size (Newell 1964), as noted here with C. asperrima. and their reproductive cycles may become less defined, occasionally exhibiting dual spawning peri- ods or continuous spawning and redevelopment (Hesselman et al. 1989, Barber and Blake 1991, Hoffmann et al. 1992). Among populations of bay scallops, Argopecten irradians. the duration of the spawning period increases with decreasing latitude (Sastry 1979). whereas mean gonad index decreases (Sastry 1970). C. asperrima are consistent with these observations to the extent that dual peaks in reproductive condition occurred in Jervis Bay in 1992 and the potential for an increase in the period of time over which spawning occurs was observed: however, a comparison with gonad index data for C, asperrima from D'Entrecasteaux Channel (Zacharin 1994) does not indicate mean gonad index changes. Differences in shell height and the availability of spawnable broodstock are of particular interest to aquaculturists and ecotox- icologists. Because C. asperrima fecundity generally increases with increasmg scallop size (Zacharin 1994, O'Connor pers. obs.). potentially greater numbers of smaller Jervis Bay scallops would be needed to produce large numbers of eggs. However, this should be overcome by the abundance of C. asperrima in Jervis Bay and the ease with which they can be induced to spawn (O'Connor and Heasman 1995). More important, the extended availability of spawnable broodstock in Jervis Bay means that embryos are also available for a longer period of time. For hatch- ery production of the species, this would extend the potential production season without the need for expensive broodstock con- ditioning. Opportunities to use C. asperrima embryos in studies of the effects of toxicants on marine organisms (Krassoi et al. in press) are also greatly enhanced. Changes in the reproductive condition of scallops have vari- ously been associated with one or more factors including latitude (MacDonald and Thompson 1988). depth (Barber et al. 1988). and genetic predisposition (Cochard and Devauchelle 1993). each of which could explain some of the differences in observations be- tween the Jervis Bay and D'Entrecasteaux Channel scallop popu- lations. In addition, temperature and food availability have been found to be of particular importance (Sastry 1968, Broom and Mason 1978. Shafee 1980) in both intcrpopulation and seasonal differences in scallop reproductive condition. Increases in Jervis Bay C. asperrima reproductive condition were coincident with decreases in water temperature and with annually recurrent phy- toplankton blooms off the NSW coast (Hallegraeff and Jeffrey 1993). These blooms can increase algal biomass 10-fold, largely as the result of short-lived diatom blooms (Hallegraeff 1981). potentially providing food for both adults and larvae. Spawning in conjunction with this annual bloom has been reported for several other molluscs on the NSW coast (Hadfield and Anderson 1988). A massive microalgal bloom of the coccolithophorid. Cephyro- capsa oceanica, occurred in Jervis Bay from mid-December to mid-January 1993 (Blackburn and Cresswell 1993). The cocco- lithophorid was readily ingested by both C. asperrima and the commercial scallop Pecten fumatiis. giving the digestive gland a distinctive milky hue. The small rise in GSI seen during this bloom suggested that it was not deleterious to scallops and that reproduc- tive condition changed in response to food availability. The dra- matic reduction in soft body weight and the rapid decline in muscle weight in December 1992. relative to 1993 (Fig. 5). may also be indicative of food availability. Changes in adductor muscle weight have been related to re- productive pattems in several scallop species. Muscle weights were found to decrease as gonad and gonadal indices increased, and vice versa (Ansell 1974, Comely 1974. Lauren 1982). In C. asperrima, there was evidence of an inverse relationship between muscle weight and the indices of gonadal development used; how- ever, no significant negative correlations were found. Peaks in muscle weight in December/January in both years corresponded to troughs in the GSI, gonad weight, and the macroscopic index. Conversely, muscle weight was at its lowest shortly before peaks in gonadal indices. In pectinids. the adductor muscle appears to be a primary site of energy storage for later use m gonadogenesis. Storage products, predominantly glycogen and some protein, are accumulated when nutrient availability exceeds net metabolic de- mand (Barber and Blake 1991). These storage products are then utilised during periods of high demand such as gametogenesis. This process can be site specific and depth dependent (Barber and Blake I99I) and. in the European scallop Pecien maximus. is controlled largely by environmental rather than genetic factors (Mackie and Ansell 1993). Although C. asperrima GSI in partic- ular showed its greatest variability during winter/spring, presum- ably as the result of gamete release, the prolonged reduction in muscle weight indicated continuing metabolic demand, consistent with continued gonadogenesis. To this extent, monitoring muscle weight m pectinids is a useful addition to tissue weight indices, particularly where partial or multiple spawnings may occur. In all weights and indices used, male and female C. asperrima varied synchronously and there was no indication of male GSI peaking earlier than that of females, as had been suggested for populations in the D'Entrecasteaux Channel (Zacharin 1994). In most collections, however, female gonads were heavier and mus- cles were lighter than those of males. Ripe females also had higher gonadal dry matter percentages and higher gross energy levels per kilogram dry weight of gonadal tissue. These weight and energy differences suggest a difference in reproductive effort between the sexes, with a 62.5-mm female having as much as 27% more en- ergy invested in the gonad. This could explain the higher average muscle weight in males, although more frequent or more intense spawning could mean that males expend similar amounts of energy over the reproductive season. In fact, observations made during this and other studies (O'Connor et al. 1994, O'Connor and Heas- man 1995) indicate that the difference in energy expended by both sexes may not be great. The scarcity of ripe males collected during this study may indicate more frequent spawning, although labo- ratory studies have indicated that male C. asperrima are more readily induced to spawn using natural cues such as temperature Reproductive Patterns in Chlamys asperrima 243 fluctuations and are capable of partial spawnings. Tiie 9% differ- ence between male and female muscle weights more likely indi- cates the difference in energy expended in reproduction, rather than a static determination of the relative energy invested in ripe gonads (279?^). Similarities in size frequency data for the sexes, and in the growth of mature scallops in previous studies (W. O'Connor pers. obs.). suggest that there is no great difference in energy expended during the reproductive period. However, in triploidA. irradians. where full maturation and spawning were prevented, muscle and soft body weights were 73 and 36% greater, respectively, than those of their diploid siblings, but shell height and length were unaffected (Tabarini 1984). Hence, the effects of differences in reproductive energy expenditure by C. asperrima are not likely to be seen in shell growth. Because C. asperrima has been harvested commercially (Young and Martin 1989) and has potential as a candidate for mariculture (Cropp 1989, O'Connor et al. 1994), variation m wet tissue weights has implications for the potential culture or harvest of C. asperrima. In markets requiring a whole or "roe on" prod- uct (muscle and gonad), scallops would be collected in winter and early spring, when the gonad is larger and more turgid. Con- versely, if muscle alone is to be sold, wet weights are greatest in summer. Differences in tissue weights between the sexes are rel- atively small in comparison to seasonal changes and are not likely to affect "roe on" sales; however, the consistently heavier male muscles would be advantageous for "roe off" markets. If C. asperrima were to be cultured, the effects of seasonal changes in reproductive condition on tissue wet weights may be mitigated by the use of triploid induction techniques, which have been found to increase muscle yield in other pectinids (Tabarini 1984). ACKNOWLEDGMENTS We thank the staff of Port Stephens Research Centre for their assistance, in particular Allen Frazer and Joseph Taylor for help in GSI determinations. Dr. John Nell and Dr. Geoff Allan, Steve McOrrie, Dan Lizska, Stephan O'Connor, and the anonymous referees for discussion or comments during the preparation of the manuscript. Thanks are also due to divers Wayne Walker and John Kelly for their help in scallop collection and to Ken O'Brien of the Wollongbar Agricultural Institute for calorimetry. LITERATURE CITED Ansell. A. D. 1974. Seasonal changes in biochemical composition of the bivalve Chlamys septemradiata from the Clyde Sea area. Mar. Biol. 25:85-95. Barber, B. J. & N, J. Blake. 1991. Reproductive physiology, pp. 377- 428. hi S. E. Shumuay (ed.l. Scallops; Biology Ecology and Aqua- culture. Elsevier, Amsterdam. Barber. B. J.. R. Gelchell. S. Shumway & D. Shick. 1988. Reduced fecundity in a deep-water population of giant scallop, Placopecleii magellanicus. in the Gulf of Maine. U.S.A. Mar. Ecol. Prog. Ser. 42:207-212. Blackburn. S. I. & G. Cresswell. 1993. A coccolithophond bloom in Jervis Bay. Australia. Aiisl. J. Mar. Freshwater Res. 44:253-260. Bricelj. V. M., J. Epp & R. E. Malouf. 1987. Intraspecific variation in reproductive and somatic growth cycles of bay scallops Argopecien irradians. Mar. Ecol. Prog. Ser. 36:123-137. Broom. M. J. & J. Mason. 1978. Growth and spawning in the pectinid Chlamys opercularis in relation to temperature and phytoplankton con- centration. Mar. Biol. 47:277-286. Chemoff. H. 1987. Factors affecting mortality of the scallop Chlamys asperrima (Lamarck) and its epizooic sponges in South Australian waters. J. E.xp. Mar. Biol. Ecol. 109:155-171. Cochard. J. C. & N, Devauchelle. 1993. Spawning, fecundity and larval survival and growth in relation to controlled conditioning in native and transplanted populations of Pecten ma.ximum (L.): evidence for the existence of separate stocks. J. E.xp. Mar. Biol. Ecol. 169:41-56. Comely, C. A. 1974. Seasonal variations in the flesh weights and bio- chemical composition of the scallop Peclen maximum (L ) in the Clyde Sea area. J. Cons. Ini. E.xplor. Mer. 35:281-295. Cropp, D. A. 1989. Scallop culture in the Pacific region, pp. 134- L53 In: M. L. C. Dredge. W. F. Zacharin & L. M. Joll (eds. ), Proceedings of the Australasian Scallop Workshop. Hobart. Tasmania. 1988. 290 pp. Daniel. C. & A. S. Wood. 1980. Fitting Equations to Data: Computer Analysis of Multifactor Data. John Wiley & Sons. New York. 458 pp. Dredge, M. L. C. 1981. Reproductive biology of the saucer scallop A/mk- sium japonicum halloti (Bemardi) in central Queensland waters. Aiisr J. Mar. Freshwater Res. 32:775-787. Gonor. J. J. 1972. Gonad growth in the sea urchin Slrongylocenlru.s pur- puratus (Stimpson) (Echinodemiata: Echinoidea). and the assumptions of the gonad index methods. J. Exp. Mar. Biol. Ecol. 12:45-64. Grant. A. & P. A. Tyler. 1983. The analysis of data in studies of inver- tebrate reproduction. I. Introduction and statistical analysis of gonad indices and maturity indices. Int. J. Invert. Reprod. 6:259-269. Grant. J. F. 1971. Scallop survey — D'Entrecasteaux Channel. Tasmania. 11-14 May. Tasmanian Fisheries Res. 5. 21 pp. Hadfield. A. J. & D. T. Anderson. 1988. Reproductive cycles of the bivalve molluscs Anadara trapezia (Deshayes). Venenipis crenata La- marck and Anomia descripta Iredale in the Sydney region. Ausl. J. Mar. Freshwater Res. 39:649-60. Hallegraeff, G. M. 1981. Seasonal study of phytoplankton pigments and species at a coastal station off Sydney: Importance of diatoms and nanoplankton. Mar. Biol. 61:107-118. Hallegraeff, G. M. & S. W. Jeffrey. 1993. Annually recurrent diatom blooms in spring along the New South Wales coast of Australia. Ausl. J. Mar. Freshwater Res. 44:325-334. Hennick. D. P. 1970. Reproductive cycle, size at maturity and sexual composition of commercially harvested weathervane scallop i^Pati- nopecten caiirinti.s) in Alaska. J. Fish. Res. Board Can. 27:2112- 2119. Hesselman. D. M.. B. J. Barber & N. J. Blake. 1989. The reproductive cycle of adult hard clams. Mercenaria spp. in the Indian River Lagoon. J. Shellfish Res. 8:4.3^9. Hoffmann, E. E., E. N. Powell, J. M. Klinck & E. A. Wilson. 1992. Modelling oyster populations III. Critical feeding periods, growth and reproduction. J. Shellfish Res. 11:399-416. Krassoi. F. R.. D. Everett & I. Anderson. In press. Methods For Estimat- ing Sublethal Toxicity of Single Compounds and Effluents to the Doughboy Scallop Chlamys asperrima {Molliisca: Pectinidae) (La- marck). National Pulp Mills Research Program Technical Report No. 18 Canberra: CSIRO. Lauren. D. J. 1982, Oogenesis and protandry in the purple hinge rock scallop Hinites giganteus. in upper Puget Sound. Washington. U.S.A. Can. J. Zool. 60:2333-2336. MacDonald. B. A. & R. J. Thompson. 1988. Intraspecific variation in growth and reproduction of latitudinally differentiated populations of the giant scallop Placopecten magellanicus . Biol. Bull. 175:361-371. Mackie. L. A. & A. D. Ansell. 1993. Differences in the reproductive ecology in natural and transplanted populations of Pecten maximus: evidence for the existence of separate stocks. J. E.xp. Mar. Biol. Ecol. 169:57-75. O'Connor. W. A . M P. Heasman. A. W. Frazer & J. J. Taylor. 1994. Hatchery rearing the doughboy scallop. Chlamys (Mimachlamys) as- 244 O'Connor and Heasman p. rimus (Lamarck), initial observations. Mem. Queensland Museum 36(2);357-360. O'Connor, W. A., & M. P. Heasman. 1995. Spawning induction and fertilisation in the doughboy scallop Chlamys {Mimachlamys) asper- rima. Aquacullure 136:117-129. Rose. R. A. & T. G. Dix. 1984. Larval and juvenile development of the doughboy scallop, Chlamys (Mimachlamys) asperrima (Lamarck) (Mollusca: Pectinidae). Aust. J. Mar. Freshwater Res. 35:315-323. Sastry. A. N. 1968. The relationships among food, temperature and gonad development of the bay xd.Wo'p^ Aequipecten irradians Lamarck. Phys. Zool. 41:44-53. Sastry. A. N. 1970. Reproductive physiological variation in latitudinally separated populations of the bay scallop, Aequipecten irradians La- marck. Bwl. Bull 138:56-65. Sastry. A. N, 1979. Pelecypoda (excluding Ostreidae). pp. 113-292. In: A. G. Giese and J. S. Pearse (eds.). Reproduction of Marine Inverte- brates. Academic Press, New York. Sause, B. L., D. Gwyther, P. J, Hanna & N. A. O'Connor. 1987. Evi- dence for the winter-spring spawning of the scallop Peclen alba in Port Phillip Bay, Victoria. Ausi. J. Mar. Freshwater Res. 38:329-337. Shafee, M. S. 1980. Application of some growth models to the black scallop Chlamys varia (L.) from Lanveoc, Bay of Brest. J. Exp. Mar. Biol. Ecol. 43:237-250. Tabarini, C. L. 1984. Induced triploidy in the Bay scallop Argopecien irradians. and its effects on growth and gametogenesis. Aquacuhure 42:151-160. Thompson, R. J. 1977. Blood chemistry, biochemical composition and the annual reproductive cycle in the giant scallop, Placopecten magellan- icus, from southeast Newfoundland. J. Fish. Res. Board Can. 34( 1 1): 2104-2116. Wells, F. E. & C. "W. Bryce. 1988. Seashells of Western Australia. West- em Australian Museum, Perth, Australia. 207 pp. West, G. J. 1990. Methods of assessing ovarian development in fishes: a review. Aust. J. Mar. Freshwater Res. 41:199-222. Young, P. C. & R. B.Martin. 1989. The scallop fisheries of Australia and their management. Crit. Rev. Aq. Sci. 1:615-638. Zacharin, W 1986. DEntrecasteaux Channel scallop survey. Tasmanian Department of Seafisheries Technical Report, No. 9. 17 pp. Zacharin, W. 1994. Reproduction and recniitment in the doughboy scal- lop, Chlamys asperrimus, in the D'Entrecasteaux Channel, Tasmania. Mem. Queensland Museum 36(2):299-306. Journal of Shellfish Research. Vol. 15, No. 2. 245-244, 1996. MONOAMINES AND PROSTAGLANDIN Ej AS INDUCERS OF THE SPAWNING OF THE SCALLOP, ARGOPECTEN PURPURATUS LAMARCK G. MARTINEZ. C. GARROTE, E. URIBE Facultad de Ciencias del Mar Universidad Caiolica del Norte P.O. Bo.\ 117 Coquimbo, Chile L. METTIFOGO, H. PEREZ, AND ABSTRACT Intragonadal injections of serotonin (5-HT). dopamine (DA), noradrenaline, and prostaglandin (PG) E; (PGE,) were assayed as inducers of spawning in the hermaphrodite scallop Argopecten purpuralus. The three monoamines were effective in inducing the release of sperm but not of oocytes. PGEj did not induce the release of either gamete. When a mixture of 5-HT or DA with PGEt was injected, gametes of both sexes were released. The injection of DA, followed (within 30 mini by an injection of PGE,. also induced the release of both gametes. This was not the case for 5-HT. The percentages of gamete fertilization and of larvae survival were much higher for the gametes spawned by an injection of DA combined with PGE, than for those gametes spawned as a result of increasing temperature and adding microalgae. These results support the hypothesis that PGs and dopaminergic mechanisms may be implicated in female spawning. This study shows that the injection of DA combined with PGE, may be successfully used in hermaphrodite scallops for obtaining viable gametes for fertilization and consequent larval development. KEY WORDS: Spawning, hermaphrodite scallops, Argopecten purpuralus, bivalve reproduction INTRODUCTION The spawning of bivalve molluscs is a process controlled by exogenous and endogenous factors (Giesc and Kanatani 1987). Among endogenous factors, prostaglandins (PGs) and somd amines, produced by nei^e cells, are considered to play an im- portant role (Khotimchenko and Deridovich 1991 , Deridovich and Reunova 1993). Gonadal dopamine (DA) content has been shown to decrease after the induced spawning of Patinopeclen ye.ssoensis (Osada et al. 1987). During the spawning of Chlamysfarreri nip- ponenesis. Matsutani ( 1990) detected an increase of the serotonin (5-HT) level in testes and ganglia, at the same time that a decrease in DA content in the ovary was observed. Martinez and Rivera (1994) reported a decrease of DA and 5-HT content in the gonads of Argopecten purpurolus during the first hours of a spontaneous spawning. A great deal of research on this subject in molluscs has referred to the induction of spawning by homogenates of nerve tissue or by monoamines, which are a common secretion product of nerve tissue (Matsutani and Nomura 1982, Matsutani and Nomura 1987, Gibbons et al. 1983, Gibbons and Castagna 1984. Braley 1985. Hirai et al. 1988. Vtjiez et al. 1990. Ram et al. 1992. Ram et al. 1993, Desrosiers and Dube 1993). 5-HT has been shown to be one of the most effective spawning inducers, although in the case of gonochoric pectinids, the sensi- tivity of males to this amine is considerably higher than that of females (Matsutani and Nomura 1982. Matsutani 1990). In the case of hermaphrodite scallops. 5-HT has been effective in induc- ing the release of sperm but not of oocytes (e.g., Argopecten irradians. Gibbons and Castagna 1984, Pecten ziczac. Velez et al. 1990). Among other monoamines assayed, noradrenaline (NA) and adrenaline were capable of inducing spawning only in male C . farreri nipponensis (Matsutani 1990). Matsutani and Nomura ( 1987) induced egg release from pieces of ovary of P. yessoensis and showed that this effect was pre- vented by the addition of aspinn (inhibitor of PG biosynthesis) and was enhanced by prostaglandin E2 (PGE,). The participation of PGs in the release of gametes has been reported by Morse et al. ( 1977) in experiments where they showed that the ultraviolet (UV) irradiation of seawater or the addition of hydrogen peroxide (ac- tivator of PG synthesis) to seawater induced spawning in male and female abalones. It has been reported that the ovarian levels of PGF,„ of Crassostrea gigas (Ono et al. 1982) and of PGE, of P. yessoensis (Mori et al. 1984, Osada et al. 1989, Osada and No- mura 1990) increased during the spawning season. It has been suggested that PGs are modulators of the 5-HT action in the in- duction of spawning in the female P yessoensis (Matsutani and Nomura 1987). This study describes the successful spawning oi A. purpuralus using a mixture of DA and PGE2 as inducers of spawning. It is also shown that the percent gamete fertilization and D-stage larval survival obtained by this method are higher than those obtained by the use of a nonchemical method. MATERIALS AND METHODS Experimental Animals Specimens of A . purpunttus were obtained from hanging cul- tures in La Herradura Bay. Coquimbo, Chile (30°S.) They were maintained in laboratory aquaria with recirculating water and fed with microalgae until they were in condition to be induced for spawning. Induction of Spawning Ripe scallops were placed in individual aquaria containing sea- water and were injected vvith 0.4 ml of the testing solution. Half of the solution was injected into the female gonadal portion, and the rest was injected into the male portion. The solutions were pre- pared by dissolving the amine in filtered seawater (FSW). Control scallops were injected with 0.4 ml of FSW. Three different ex- periments were done: in the first one, animals were injected with only one compound, in the second experiment, one monoamine and PGE2 were injected together as a mixture; and in the third 245 246 Martinez et al. expi. .ment. the amines and the PG were both injected in the same animal , but separately, one 30 min after the first one had been injected. The first and the third experiments were assayed twice each, and the second experiment was assayed three times. Each assay was conducted on a different date, but each time, a set of animals was obtained from the same population and randomly divided into experimental and control groups. The response to the testing solution was recorded as positive when gametes were re- leased. Statistical differences among experimental and control re- sponses were analyzed by the use of the Fisher exact probability test. Fertilization and Larvae Survival An experiment was designed to compare percent fertilization and survival of D-stage larvae obtained when spawning was indi- vidually induced by a mixture of DA and PGE, and when spawn- ing was induced by a method that consisted of adding microalgae and increasing temperature for groups of scallops placed in the same tank (referred to as usual method). In both cases, when scallops started to release sperm, they were removed from the aquarium, rinsed with FSW, and placed in another tank containing FSW. Sperm obtained by each of the two methods was mixed separately and maintained for later fertilization. After a few spawning contractions, scallops were individually transferred to a third clean tank, where they remained until oocyte release began. When this occurred, the scallops were rinsed with clean FSW and placed into individual plastic containers with FSW, where they continued to spawn. When spawning seemed to have ended, the scallops were removed from the containers and the number of oocytes in each container was estimated by the counting of repli- cate 1-ml samples. Fertilization was initiated by the addition of spermatozoa to individual oocyte suspensions. The final ratio of spermatozoa to oocytes was 10:1. Percent Fertilization Percent fertilization was calculated as the ratio of the number of dividing oocytes (embryonic state) and the total number of oocytes (fertilized and unfertilized). 2 h after the addition of spermatozoa to oocytes. Samples of 1 ml of suspension from each culture were taken, fixed with I, -iodine solution, and kept for later counting of dividing embryos and intact oocytes. D-Stage Larval Survival The rest of the suspension cultures was combined according to the spawning method and placed in four 50-1 tanks containing FSW at final densities of 30 eggs/ml. They were left undisturbed until they developed to D-stage larvae (about 48 h later). The larvae were then carefully filtered, rinsed, and resuspended in FSW. Five I ml samples were taken from each suspension culture, fixed with U-iodine solution, and kept for later counting of D-stage larvae. 1 his experiment was replicated in two different periods of the year, once in December (summer time) and again in August (win- ter time). Results were analyzed and are presented separately. After arc-sin transformations, percent fertilization and D-stage lar- val survival were compared by a simple analysis of variance. RESULTS Induction to Spawning Assay of Compounds Alone 5-HT, injected at doses of 2 ■ 10*^ and 2 • 10"^^ M, induced the release of sperm in 100% of the scallops, but only one animal from each treatment released oocytes (Table 1). At the lowest dose, none of the animals injected with DA released either sperm or oocytes. At the highest dose, 100% of the scallops released sperm, but only one scallop from the first assay released oocytes. Injections of NA at a dose of 2 • 10 ~' M, except for one individual that released sperm, did not induce the release of ga- metes of either sex. When NA was injected at a higher dose (2 • 10 "-^ M), 90% of the scallops released sperm and one scallop from each assay released oocytes. PGEn, in all doses tested, did not induce the release of gametes in either sex. Control animals injected with FSW alone did not induce the release of gametes in either sex. Assays of Monoamines Combined With PGE2 When a mixture of 5-HT (2 • 10" '' M) and PGE, (2 • 10^* M) was injected, 100% of the animals responded by ejecting sperm and 41% of them responded by releasing oocytes (Table 2). DA (2 • 10 ' M) combined with PGE, (2 • 10^* M) induced the release of sperm in 100% of the scallops injected and the release of oocytes in 41% of them. The mixture of NA (2 • 10 """M) and PGE, (2 • 10"^ M) induced the release of sperm in 82% of ani- mals injected and the release of oocytes in 18% of them. The total number of animals (assays 1, 2, and 3) that released oocytes after injections of 5-HT and PGE, and injections of DA and PGE-, was statistically significant (p < 0.05. two-tailed Fisher exact proba- bility test). Assays of Monoamines and PGEj Injected Separately When DA or 5-HT was injected before or 30 min after an injection of PGE,. 100% of the scallops released sperm (Table 3). Only two animals (out of 10) released oocytes when DA was injected after PGE,, and 407f of them (p < 0.05, one-tailed exact probability test) did so when the order of injection was reversed TABLE 1, Induction of gamete release by monoamine or PGEj injection in the scallop, .4. purpuratus. ig Solution First Assay Second Assay and Concentration (M) Oocytes Sperm Oocytes Sperm 5-HT 2 X 10--' 0/5 5/5 1/5 5/5 5-HT 2 X \Q-^ 1/5 5/5 0/5 5/5 DA 2 X 10"' 0/5 0/5 0/5 0/5 DA 2 X 10'^ 1/5 5/5 0/5 5/5 NA 2 X 10"^ 0/5 0/5 0/0 1/5 NA 2 X 10-' 1/5 4/5 1/5 5/5 PGE, 2 X 10"* 0/5 0/5 0/5 0/5 PGE, 2 X lO""* 0/5 0/5 0/5 0/5 FSW 0/5 0/5 0/5 0/5 Results are expressed as number of animals that released gametes/number of animals tested. Monoamines and PGE^ as Spawning Inducers 247 TABLE 2. Induction of gamete release by an injection of a monoamine combined with PGE, in the scallop, A. purpuratus. Assay 1 Assay 2 Mixtures* Oocytes Sperm Oocytes Sperm Assay 3 Oocytes Sperm PGE; + 5-HT PGE, + DA PGE, + NA FSW 3/5 2/5 1/5 0/5 5/5 5/5 2/5 0/5 0/7 2/7 2/7 0/7 7/7 7/7 7/7 0/7 4/5 3/5 0/5 0/5 5/5 5/5 5/5 1/5 Results are expressed as number of animals that released gametes/number of animals tested. * PGE, dose. 2 x 10"" M; 5-HT. DA. or NA doses. 2 x 10"-' M. (first DA ;md then PGE,). In none of the three assays did FSW injections (control) alone induce spawnmg. Fertilization and Larvae Survival A higher proportion of oocytes was fertilized when gametes were obtained by the injection of 0.4 ml of 2 • 10" ' M DA mixed with 2 • 10"'' M PGE,, than when they were obtained by adding microalgae and increasing temperature (Table 4). Less than 50% of D-stage larvae obtained from the fertilization of gametes re- leased by adding microalgae and increasing temperature survived, whereas inore than 617c of those obtained by the injection of DA and PGE, survived (Table 4). Similar results were obtained for experiments conducted during both summer and winter periods. DISCUSSION Results obtained with injections of 5-HT on the spawning of A. purpuratus were similar to those obtained by Gibbons and Cast- agna ( 1984) and Velez et al. (1990) for the hermaphrodite species A. irradians and P. ziczac. In all of these cases, it was shown that injections of 5-HT were effective at inducing the release of sperm but not oocytes. In the case of gonochoric bivalves, most of the studies have shown that the sensitivity of male individuals to 5-HT induction is higher than that of females. This higher sensitivity may be expressed either as a lower dose of the amine necessary to induce the release of gametes (Matsutani and Nomura 1982. Mat- sutani, 1990) or as a higher percentage of animals that spawn following 5-HT injection (Gibbons et al. 1983. Gibbons and Cast- agna 1984, Braley 1985. Belda and Del Norte 1988). Except for a few successful assays on inducing the release of TABLE 3. Induction of gamete release by monoamine and PGE, injected separately in the scallop, .4. purpuratus. First Experiment Second Experiment Testing Solution" Oocytes Sperm Oocytes Sperm PGEj ^ 5-HT 0/5 5/5 0/5 5/5 PGE, -* DA 0/5 5/5 2/5 5/5 5-HT -^ PGE, 0/5 5/5 0/5 5/5 DA -^ PGE, 3/5 5/5 1/5 5/5 FSW -^ FSW 0/5 0/5 0/5 0/5 Results are expressed as number of animaK that released gametes/number of animals tested. * PGE, dose, 2 X 10"" M; 5-HT. DA. or NA doses, 2 x 10"' M. Arrows indicate that the second compound was injected 30 min after the first one. sperm (Hiraietal. 1988, Matsutani 1990). injections of DA or N A have not been demonstrated to be good inducers of spawning. Matsutani and Nomura (1986) have suggested that UV-irradiated seawater and rising temperature induce female spawning by stim- ulating serotonergic mechanisms via dopaminergic mechanisms. When DA or NA was injected as a single compound (these re- sults), only sperm was released by the scallops. However, the injection of a mixture of any of these amines or 5-HT with PGE, did induce oocyte release in ,4. purpuratus. Whenever DA was injected with PGE,, either in combination or before it. the release of oocytes was shown. This result was not obtained for 5-HT. These findings are consistent with a suggested (Osada et al. 1989. Matsutani 1990) difference in spawning mechanism be- tween female and male scallops. We have studied (Martinez et al. 1996) changes in the levels of DA. NA. and 5-HT in separate ganglia oi A. purpuratus associated with non-drug-induced spawn- ing. In those individuals that spawned. DA and NA decreased in the visceral ganglion (innervating mainly the female gonadal por- tion) and did not change in the cerebropedal ganglion (innervating mainly the male gonadal portion). On the other hand, spawning was associated with a decrease in 5-HT in the cerebropedal gan- glion and no change in the visceral one. Matsutani and Nomura ( 1987) have suggested that PGs may be modulators of 5-HT action in the female Japanese scallop P. yes- suensis. They showed that when aspirin was present. 5-HT- induced egg release was inhibited, whereas it was enhanced by PGE,. Morse et al. ( 1977) have reported the spawning oi Haliotis rufescens induced by the addition of PGE, PGF,„ or by hydrogen TABLE 4. Percentage of gamete fertilization and of D-stage larval survival of A. purpuratus obtained by two different methods of induction of spayvning. Usual Induction of Spayvning (%)* Chemical Induction of Spawning (%)t Gamete fenilization Summer experiment Winter e.xperiment Larvae D survival Summer experiment Winter experiment 52.8 ± 6.4 62,8 ± 2.9 30.3 ±8.0 29.2 ± 6.7 88.4 ± 2.2** 84.0 ± 0.1** 67.5 ± 10.7** 70.8 ± 5.9** Results of two experiments are presented as the mean ± SD (n summer experiment and n = 3 for winter experiment). * Addition of microalgae and increasing of temperature (usual). ** Significantly different (p < 0.01). t Injection of a mixture of PGE, and DA Uhenucal). 5 for 248 Martinez et al. pert .ide (activator of PG synthesis) to seawater. Matsutani and Nomura (1986) did not find any effect of PGF2„ on tlie induction of spawning in P. yessoensis. Osada et al. ( 1989) showed that the levels of PGE2 and PGFic were about four times higher in the ovary than in the testis of P. yessoensis and that during spawning, these values increased in the male gonad whereas they decreased in the female one. Ram et al. (1992) have proposed a model for the regulation of spawning in bivalves. They propose that males and females detect environmental cues by specific chemical sensing receptors and that signals are conducted by the nervous system to the gonad, where they may directly activate it or may induce the release of another hormone that activates the release of gametes. They propose that 5-HT might be the intermediate substance between the nervous system and gonads, but its action is modulated or mediated by PGs. The use of DA combined with PG may be a very good method to induce spawning, al least for hermaphrodite scallops, because of the high percent fertilization and survival of the resulting larvae, both of which were much higher than results obtained when we used the method of increasing temperature or adding microalgae. As far as we know, this is the first comparative analysis of spawn- ing induction methods using embryos and larvae as the focus rather than the actual yield of spawning. Belda and Del Norte (1988) showed nearly 100% fertilization of oocytes stripped from the ovaries of the scallop Amusiuin pleiironectes when using sperm induced by 5-HT, but the development of fertilized eggs into lar- vae on the first day was only 0.739^. About I07f survival of veligers from fertilized eggs of the china clam, Hippopus porcel- lanus, was obtained when specimens were induced to spawn by the injection of 5-HT (Alcazar et al. 1987). We think that when spawning is induced by the mixture of DA and PGE,, most of the gametes that are released are those in the best state of ripeness, and this would not be the case with other induction methods. In this late case, the release of gametes might be the result of a mechan- ical stimulation (muscular contraction). We have observed that with this mixture (DA and PGE^) the gonad does not look quite as empty after the release of oocytes as when spawning is induced by other factors. Ram et al. (1993) have reported that in the zebra mussel, Dreissena polymorpha, the likehood of female spawning in response to 5-HT was not tightly coupled to the morphological maturity of the gonad, and they conclude that another maturational process in the gonad must be necessary for oocyte release. These results suggest that the release of oocytes in the hermaphrodite scallop A. purpuralus is controlled by dopaminergic pathways under the regulation of PGs. ACKNOWLEDGMENTS We thank Mr. Raul Vera, Carlos Solar, and the staff of the Unidad de Produccion of Facultad de Ciencias del Mar. Univer- sidad Catolica del Norte, for their technical assistance. We are very grateful to Dr. Raymond Bienert for his help with English language. This work was supported by the Fondo Nacional de Investigacion Cientifica y Tecnologica (FONDECYT GRANT # 194-1125). LITERATURE CITED Alcazar, S. N., E. P. Solis & A. C, Alcala. 1987. Serotonin-induced spawning and larval rearing of the china clam, Hippopus porcellcinus Rosewater (Bivalvia: Tridacnidae). Acjuaciilnire 66:359- .^6X. Belda, C. A. & A. G. C. Del Norte. 1988. Notes on the induced spawning and larval rearing of the Asian Moon Scallop, Amusium pleiironecles (Linne), in the laboratory. Aqiiaculnire 72;173-I79. Braley, R. D. 1985. Serotonin-induced spawning in giant clams (Bival- via:Tridacnidae). Aquacullure 47:321-325. Deridovich, I. I. & O. V. Reunova. 1993. Prostaglandins: reproduction control in bivalve molluscs. Camp. Biochem. Physiol. 104A:23-27. Desrosiers, R. R. & F. Dube. 1993. Flowing seawater as an inducer of spawning in the sea scallop, Placopecten magellanicus (Gmelin, 1971). y. Shellfish Res. 12:263-265. Gibbons, M. C, J. G. Goodsell. M. Castagna & R. A. Lutz. 1983. Chemical induction of spawning by serotonin in the ocean quahog Arlica islandica (Linne), J. Shellfish Res 3:203-205. Gibbons, M. C. & M. Castagna. 1984. Serotonin as an inducer of spawn- ing in six bivalve species. Aquiiculture 40:189-191. Giese, A. C. & H. Kanatani. 1987. Maturation and Spawning, pp. 251- 329. In: A. C. Giese, J. S. Pearse and V. B. Pearse (eds). Reproduc- tion of Marine Invertebrates, vol. 9. Blackwell Scientific Publications. Palo Alto, California. Hirai, S., T. Kishimoto. A. L. Kadam. H Kanatani & S. S. Koide. 1988. Induction of spawning and oocyte maturation hy 5-hydroxytryptamine in the surf clam. J. Exp. Zool. 245:318-321. KhoiiTichenko, Y. S. & I. 1. Deridovich. 1991. Monoaminergic and cholinergic mechanisms of reproduction control in marine bivalve mol- luscs and echinoderms: a review. Comp. Biochem. Physiol. I00C:31 1- 317. Martinez, G. & A, Rivera. 1994. Role of monoamines in the reproductive proceis of Aii>opeilen piirpiiriilHS. Imcrtclv. Reprod. Dew 25(2):167- 174. Martinez, G., F. Saleh, L. Mettifogo, E. Campos & N. Inestrosa. 1996. Monoamines and the release of gametes by the scallop Argopecten purpiiratus. J. Exp. Zool. (In press!. Matsutani. T. 1990. Endogenous factors controlling spawning in marine bivalves, pp. 231-237. In: M. Hoshi and O. Yamashita (eds.). Ad- vances in Invertebrate Reproduction, vol. 5. Elsevier Science, Amster- dam Matsutani. T. & T. Nomura. 1982. Induction of spawning by serotonin in the scallop Pulinopecten yessoensis (Jay). Mar. Biol. Lett. 3:353-358. Matsutani. T. & T. Nomura. 1986. Pharmacological observations on the mechanism of spawning in the scallop Piitinopeaeii yessoensis. Bull. Jpn. Soc. Sei. Fish. 52:1589-1594, Matsutani. T. & T. Nomura. 1987. In viiro effects of serotonin and pros- taglandins on the release of eggs from the ovary of the scallop Pali- nopecten yessoensis. Gen. Com. Endocr. 67:111-118. Mon. K.. M, Takenaga. T, lino & T. Nomura. 1984, Prostaglandin Fjc profile during the spawning season in the scallop, Pulinopecten yes- soensis. Kyoto Conference on Prostaglandins, Abstracts, p, 250, Kyo- to. Nov. 1984. Morse. D. E.. H. Duncan. N. Hooker & A. Morse. 1977 Hydrogen peroxide induces spawning in molluscs, with activation of prostaglan- din endoperoxide synthetase. Science 196:298-300, Ono. K,. M. Osada. T. Matsutani. K. Mori & T. Nomura. 1982. Gonadal prostaglandin F,„ profile during sexual maturation in the oyster. Cras- sostrea gigas Thunberg. Mar. Biol. Lett. 3:223-230. Osada. M.. T. Matsutani & T. Nomura. 1987. Implication of catechol- amines during spawning in manne bivalve molluscs. Invertebr. Re- prod. Dev. 12:241-252. Monoamines and PGE, as Spawning Inducers 249 Osada. M.. M. Nishikawa & T. Nomura. 1989. Involvement of pros- taglandins in the spawning of the scallop. Patinopeclen yessoensis. Comp. Biochem. Physiol. 94C;595-601. Osada. M. & T. Nomura. 1990. The lesels of prostaglandins associated with the reproductive cycle of the scallop. Patinopeclen yessoensis. Prostaglamlins 40:229-239. Ram. J L.. P. Fong. R. P. Croll. S. J. Nichols & D. Wall. 1992 The zebra mussel (Dreissena potymorpha). a new pest in North America: reproductive mechanisms as possible targets of control strategies, /n- verlebr. Reprod. Dev. 22:77-86. Ram. J. L., G. W. Crawford. J. U. Walker, J. J. Mojares, N. Patel, P. P. Fong & K. Kyozuka. 1993. Spawning in the zebra mussel (Dreissena polymorpha): activation by internal or external application of seroto- nin. J. Exp. Zool. 265:587-598. Velez. A.. A. Alifa & O. Azuaje. 1990. Induction of spawning by tem- perature and serotonin in the hermaphroditic tropical scallop. Pecten ziczac. Aquacullure 84:307-313. Journal of Shellfish Research. Vol. 15. No 2. 251-257, 1996. VELIGERS FROM TWO POPULATIONS OF SCALLOP PLACOPECTEN MAGELLANICUS EXHIBIT DIFFERENT VERTICAL DISTRIBUTIONS IN THE SAME MESOCOSM* JOAN L. MANUEL,' SUSAN BURBRIDGE,' ELLEN L. KENCHINGTON.- MARTIN BALL,' AND RONALD K. ODOR' 'Biology Department Dalhousie University Halifax, Nova Scotia. Canada BJH 4J1 'Invertebrate Fisheries Division Science Branch Department of Fisheries and Oceans P.O. Box 550, Halifax. Nova Scotia. Canada B3J 2S7 ABSTRACT Veligers of the giant scallop. Placopecien magellaniciis. spawned from parents that came from two different popula- tions (Georges Bank and Passamaquoddy Bay), were maintained together in large mesocosms 0.6 m in diameter and 9.0 m deep. A new microsatellite probe (PMMS-130) developed by Gjetvaj et al. (unpublished observation) was used to distinguish the population of origin of the veligers. Samples obtained at various depths confirmed that the vertical distribution of the two populations differed. These results confirm population differences in vertical migration behavior seen in a previous study and indicate the value of the microsatellite probe for larvae as well as the efficacy of mesocosm replication techniques. KEY WORDS: Scallop, larvae, genetic. DNA, migration INTRODUCTION Several field studies have found diel changes in the vertical distribution of bivalve veligers (Maru et al. 1972, Harding et al. 1986. Scrope-Howe and Jones 1986. Tremblay and Sinclair 1990, Raby et al. 1994). Mesocosm studies isolate the effect of the behavior of experimental organisms from that of abiotic transport and have demonstrated that veligers of both the great scallop {Pecten maximus) and the giant scallop (Placopecten magellani- ciis) do make distinct vertical migrations (Kaartvedt et al. 1987, Silva-Serra and 0"Dor 1994, Gallager et al. 1996. Manuel 1996). Although much useful knowledge has been gained through con- trolled mesocosm experiments, such studies have traditionally been hampered by a reluctance to remove too many animals (which leads to low numbers and a loss of statistical robustness), a lack of replication and control in experimental manipulation, and contamination by other species (Balch et al. 1978). Recently, Manuel (1996) solved several of these problems by using noninvasive sampling (video profiling) and replicated me- socosms to describe interpopulation differences in the vertical dis- tributions of P. magellanicus veligers. The differences in vertical distribution were due to differences in veliger migration behavior and in veliger responses to discontinuities such as the surface and a thermocline placed middepth in the mesocosms. Although the replication of mesocosms allows statistical testing, the number of mesocosms (and thus the power of statistical tests) was constrained by logistics. Variability among mesocosms, as reflected by mean depths (Fig. 1). populations of volunteer organisms, and veliger growth rates, was high (Manuel 1996). This was because repli- cated mesocosms are not simply replicated physical media, but 'Contribution to the program of OPEN (Ocean Production Enhancement Network, one of the 15 Networks of Centres of Excellence supported by the Government of Canada) and IFRP (Interim Funding Research Pro- gram) rather are replicated ecosystems that may differ considerably by the end of a long experiment. Thus, some doubt remains as to whether the differences in vertical distribution observed among populations are the result of different responses to the same stimuli or different responses to different stimuli provided by a different community of organisms. Even if those community differences were caused in some way by the different populations (which would lead to the same statistical result), the results of Manuel (1996) would be biologically insignificant. Concentrations of P. magellanicus veligers are unlikely to be high enough, in an open natural system, to make substantial ecosystem changes. Even with a light microscope, many bivalve species are indis- tinguishable as veligers (Hurley et al. 1987. Demers et al. 1993). This problem has been overcome by recent advances in scallop molecular biology. Gjetvaj et al. (in prep.) have developed a num- ber of microsatellite probes that can be used (with polymerase chain reaction amplification) for both adult scallops and individual veligers. Microsatellites are permutations of simple repeated se- quences of DNA up to a few hundred base pairs in length dispersed throughout the genome. Microsatellite alleles vary in length (i.e., in the number of bases) and exhibit high levels of polymorphism, which make them ideal for use as genetic markers. This experi- ment is the first to use microsatellite loci to identify the parents of individual scallop veligers. We used the technique to identify the population of origin of scallop veligers at different depths within the same mesocosm. We were able to demonstrate that veligers of P. magellanicus. spawned from adults from Passamaquoddy Bay. are found shallower in the same mesocosm than those spawned from adults from Georges Bank. This experiment addresses the potential problems with small numbers of replicates in variable mesocosms. tests the practical potential of the new genetic probe developed by Gjetvaj et al. (in prep.), and confirms the results of a previous experiment (Manuel 1996) with different parents and conditioning methods. 251 252 Manuel et al. 16:00 16:00 4:00 Time (h) I6O0 Age (d) Figure 1. Top two panels after Manuel (1996): ZCM, recorded at 2-h intervals at the age of 28-31 d. Each line plots the ZCM of a single mesocosm over 2.5 d. Clear boxes indicate when lights were on, and black boxes indicate darkness. Bottom panel: comparison of the nom- inal depth (average of two replicates except at age 14 d) of GEO and PAS veligers in this experiment. METHODS Conditioning and Spawning of Adults Forty-four adult P. magellanicus from the Georges Bank (GEO) and Passamaquoddy Bay (PAS) populations were held at a Mahone Bay aquaculture site for 6 months. These were brought to the Aquatron facility at Dalhousie University and stimulated to spawn in September 1993. Scallops were genotyped (see Micro- satellite Techniques below) with a microsatellite probe (PMMS- 130) developed by Gjetvaj et al. (unpub. obs.). Because some alleles were found in both populations, parents for the experiment were chosen such that each allele appeared in only one or the other population. Unfortunately, misreading of the score of one individ- ual resulted in one allele (134) being represented in both popula- tions (Table I ). However, because each veliger was represented by two alleles, this did not cause any serious difficulties, and we were able to identify the population of origin of individual veligers by determining the genotype of each veliger. We crossed only within the same population for this initial experiment. After the gonads of the scallops had been emptied, the selected parents were reconditioned by supplying them with ad libitum Isocluisis galbcuia (clone TISO) from October I. 1993, until spawning on February 27. 1994. Ripe animals were cleaned with a scrubbing brush in clean filtered seawater and induced to spawn in clean. 0.2-(i,m-pore-size filtered seawater by thermal stimula- tion and agitation of the water with a submersible pump. Both populations were spawned concurrently, but in separate rooms, with a different person handling each population. Hot fresh water was used frequently to rinse hands and minimize cross- contamination. The eggs from all females in a given population were mixed together and fertilized with the mixed spawn of all of the males from that population. Sperm was added until several sperm could be seen around each egg when a sample was exam- ined under a light microscope. Fertilized embryos were introduced to the surface of 9-m-deep mesocosms filled with l-)xm-pore-size filtered seawater as soon as fertilization success had been assured (i.e.. when most of the embryos were in cleavage or blastula stage). Each polyethylene mesocosm in the 10-m tower tank was tied at the bottom with a watertight knot, suspended from the surface with a styrofoam collar, and filled slightly above the level of water in the tower tank to create positive pressure. Each pop- ulation was held, unfed, in separate mesocosms until 4 d of age (D-stage). The process of filling the experimental mesocosms modified the preexisting temperature regime somewhat, resulting in temperatures above the thermocline between 14 and 16°C and below the thermocline between 6 and 10°C over the 4 d of devel- opment from egg to D-stage veligers (Fig. 2A). Treatment of Veligers At 4 d. veligers were concentrated on a 53-ji,m nitex screen, counted, and distributed to two experimental mesocosms. Our objective was to stock each mesocosm with 5.0 x 10' veligers TABLE 1. Microsatellite scores of parental alleles from the GEO scallops and number of progeny produced from each of the crosses. One male showed three instead of two bands. The 126 and 162 bands in male G22M segregated together. George: Bank Female GllF Female G3F Total Scort 144 136 146 140 By allele By scallop Male G22M 162/126 i.;6 53 73 58 56 37 35 25 24 173 188 361 Male G37M 154 24 24 19 15 82 154 150 26 19 14 13 72 Total (by allele) 176 157 105 77 total (Georges Bank) Total (by scallop) 333 182 515 Veligers of p. Magellanicus 253 u 1 1 1 1 r A ■ T ■ • - - - ; _ 1 ^ S & 0 Q ^^ ' " 1 ' \ - \ ■ 9 in .■ 1 1 1 1 1 1 1 1 4 6 8 10 12 14 16 18 20 4 6 8 10 12 14 16 18 20 Temperature °C Temperature °C Figure 2. Temperature in the tower tank. (A) Temperature during development to D-stage. Solid line is day 2, broken line is day 3, and dotted line is day 4. (B) Temperature in the tower tank on sampling days. Solid line is day 14, large broken line is day 19, small broken line is day 33, and dotted line is day 41 . The nominal depth of veligers from each sample day is represented by circles (PAS) and diamonds (GEO). Shades symbols are day samples; black symbols are from the night sample. from each population. However, as with a previous experiment (Manuel 1996). we found that GEO veligers developed slightly more slowly than PAS veligers. We did not want to leave PAS veligers unfed for an extra day while we waited for all GEO veligers to get to D-stage and were unsure about the viability of GEO larvae not quite in D-stage after they had been captured on a screen. We decided to redistribute veligers as planned on the fourth day and to exclude individuals not completely in D-stage from sample counts when stocking the mesocosms. Gentle han- dling apparently allowed most of the trochophores to survive the process, resulting in more GEO than PAS veligers in each meso- cosm (about a 60/40 split). Mesocosms were also established that contained veligers from only one of the populations (four meso- cosms each for GEO and PAS) to monitor growth rates of the two populations. Treatment mesocosms were filled with 0. l-(i,m-pore-size fil- tered seawater and inoculated with enough cultured TISO to bring the concentration to 1 .0 x 10"* cells • ml ' . A gravity-fed perfo- rated vinyl sprinkler hose was used to evenly distribute food from the top to the bottom of the mesocosm. At 24 d of age. veligers from both mesocosms were concentrated on a 80- (xm nitex screen, sampled, mixed thoroughly, and evenly redistributed in two clean mesocosms. The particle level in the mesocosms was monitored throughout the experiment with a Multisizer Coulter Counter*, and supplemental TISO was added whenever mean concentrations were near or below 5.0 x 10' cells • ml '. Supplemental TISO was added on nine occasions (days 6, 12, 14, 17, 19, 25, 33, 35, and 39), resulting in particle levels that varied between 4.0 x 10^ and 1 .40 x 10"* cells • ml ' through the experiment. Stratification Heating and cooling coils in the tower tank were set at 6-m depth, and we attempted to maintain a 5°C thermocline over 0.5 m. The choice of thermocline strength was relatively arbitrary: because we had a strong thermocline ( IO°C over 0.5 m) in our first experiment (Gallager et al. 1996, Pearce et al. 1996) and a very weak thermocline (1.5°C over 0.5 m) in our second experiment (Manuel 1996), this experiment was meant to be midway between the two. Samples for Genotyping of Veligers Samples of veligers were taken from the surface by dipping with a beaker and at depth by siphoning with a garden hose. The siphon was started in filtered seawater, paused, and then lowered sequentially to the appropriate depths for samples. At each depth, a volume slightly greater than the contents of the siphon hose was drawn before the sample was taken, to prevent contamination by veligers from other depths. Tests of the procedure at the beginning of the experiment using water with and without veligers showed that veligers were flushed from the hose in the parcel of water in which they entered and that there was no apparent mixing in the siphon, even if the siphon was paused for several minutes. Sam- ples were collected from each mesocosm at the surface (0 m), the middle of the layer above the thermocline (2.5 m). just above the thermocline (5 m). and from below the thermocline (8 m). We removed water until we had 30 individuals for each sample, except that if we had not collected enough veligers after 20 1 had been removed for any given sample, we stopped (concentrations of veligers were often very low below the thermocline). Water re- moved in sampling was replaced with fresh filtered seawater. The technique used to collect veligers from the mesocosms for geno- typing was much less sensitive for resolving the depth of veligers than the video sampling, so we chose to sample several times and to combine the results, to avoid the chance of accidentally choos- ing the one time when the two populations did not differ substan- tially. Samples for genotyping were collected from each meso- cosm on six occasions: four times midday (ages 14. 19. 33. and 42 d). once at night (age 41 d), and once when the mesocosms were changed midexperiment (age 24 d). Only the 24-d sample was not depth stratified. That sample was collected after the veligers had been concentrated on a screen to confirm the relative proportions of veligers from each population in the mesocosms (we considered the possibility that the ratio was affected by the differences in distributions). The data from one sample (mesocosm #2 at age 14 d) were discarded because of an obvious recording error. We had data for veligers below the thermocline when records show that none were collected at that depth on that day. and two few veligers were recorded from the first two depths. Samples thus collected were taken immediately back to the laboratory and narcotized by the addition of an equal volume of 1 M Tris-EDTA to the sample. Thirty individual veligers were ex- tracted with a micropipette in 5 (xl of the solution and frozen in separate microcentrifuge tubes at -60°C. Microsatellite Techniques DNA was collected with a syringe to extract approximately I ml of haemolymph from the adductor muscle sinus of each pro- 254 Manuel et al. spe: ave parent. The haemolymph was then added to an equal \ lume of 95% ethanol. A 200-|jil aliquot of each ethanol- preserved blood sample was centrifuged for 2 min at 2.000 rpm. The pellet was resuspended in a TE buffer ( 10 mM Tris [pH 7.5], 1 mM EDTA) and repelleted by again centrifuging for 2 min at 2,000 rpm. The pellet was then suspended in a high-salt lysis buffer (400 mM NaCl, 10 mM Tris (pH 8.2], 10 mM Na EDTA). Triton 100 (0.75%) and 500 jjig • ml" ' proteinase K were added to the lysate and incubated overnight at 55°C (see Patwary et al. 1994). The protein present in the lysate was precipitated with 400 |xl of saturated NaCI solution and vortexed vigorously for 15 min. The tubes were then centrifuged for 30 min at 800 rpm to pellet the precipitated proteinaceous debris. The nucleic acids present in the supernatant were further purified with an equal volume of chloro- form. The DNA was then precipitated from the aqueous layer with equal volumes of cold isopropanol. The sample was finally pel- leted at 14,000 rpm for 30 min at 4°C. dried, and resuspended in the TE buffer. DNA was extracted from the veligers by digesting the entire animal in 50 |xl of a solution containing 20 mg • ml ~ ' proteinase K. 0.5% Tween 20, 50 mM KCl, 10 mM Tris-HCl (pH 8.5), for 2.5 h at 55°C. The solution was then incubated at 95°C for 5 min and stored at 4°C until used. Polymerase chain reaction (PCR) amplification of the scallop microsatellite locus PMMS-130 was performed in a 20-|jl1 reaction mix containing 1-2 jjlI of template DNA of unknown concentration. 50 mM KCl, 10 mM Tris-HCl (pH 8.5), 1 mMMgCl,,0.5 (jlM each of primer PMMS-130F and PMMS-130R, approximately 1 U of Taq polymerase and 0.2 mM each of ATP, GTP, CTP, and TTP. The reaction mixture was subjected to 25 cycles of 20 s at 94°C, 20 s at 49°C. and 20 s at 72°C to allow the PCR to proceed. The DNA from 2.5 jjlI of each PCR amplification mix was size separated on an 8% acrylamide denaturing, sequencing gel accord- ing to the directions provided in the Pharmacia T7 sequencing kit. Each gel contained as a size standard several sequencing reactions of the M13 single-strand DNA template provided in the Pharmacia sequencing kit. The sizes of the microsatellite alleles were deter- mined by comparing the migration distances of the PCR products with the migration distances of the M13 standards. Scores indi- cating uneven numbers of base pairs were taken as the next higher even number. Two independent readers viewed each gel, and where results differed, the gels were reexamined. If a consensus could not be reached, the amplified product was size separated again. Because all alleles except 154 were represented in only one population each, errors in reading, chemistry, two veligers in one sample, or unintentional crosses appeared as impossible combina- tions from the parents, and those veligers could not be assigned to either population. If this occurred, the datum was rejected. Statistical Analysis Differences in distribution were determined by x' analysis with a significance level of p = 0.05 and Bonferroni testing where multiple tests were used (Sokal and Rohlf 1981). Because mean depth (ZCM) is affected by several factors other than simply the time of day (Manuel 1996), raw data from different days and times could not be simply combined to test the overall results. The significance of the experiment as a whole was determined by Sokal and Rohlf's (1981) method for combining probabilities from in- dependent tests of significance (pp. 779-782). For the purpose of determining whether GEO veligers were deeper than PAS veligers in the mesocosms, we calculated the nominal depth (d„) for each population; where n, is the number if individuals in that population at depth d,, and np is the total number of veligers from that population from all depths on thai date. Because we sampled an equal number of veligers, rather than an equal volume of water at each depth, the nominal depth describes the depth of the veligers relative to the other population, rather than the mean depth of the population. In other words, if one population is deeper than the other, its nominal depth calculated in this way will be deeper, but the nominal depth is not the equivalent of mean depth (ZCM) in earlier reports. RESULTS Instabilities in the temperature control system made it neces- sary to manipulate the thermal controls to maintain the desired degree of stratification (5°C). On sampling days, the degree of stratification ranged from 3.7 to 10.3°C, and neither the nominal depth nor the differences between the populations were correlated with the degree of stratification (Fig. 2B). Although we were unable to identify the population of veligers for the purposes of size measurements, veligers raised from the same spawning of these two populations in separate mesocosms (four mesocosms for each population) did not show any significant differences in growth rates between the two populations (Fig. 3). All possible a 60 a V a CO u Age (d) Figure 3. Mean size of veligers in mesocosms containing only PAS (solid line) or GEO (dotted line) veligers. Error bars are SD, n = 4 for each population. Veligers have the same parents and were raised in the tower tank at the same time as mesocosms containing veligers from both populations. Veligers of p. Magellanicus 255 TABLE 2. Microsatellite scores of parental alleles from the PAS scallops and number of progeny produced from each of the crosses. The 134 x 152 cross (*) could have been produced by three different crosses, and the 152 x 152 cross (**) could have been produced by two crosses, so the total number for each of these is listed only once. Passamaquoddy Bay Female P4F Female PIOF Female PIF Female P8F Score 168 134 138 152 152 142 154 158 Male P18M 134 11 20 25 87* * 17 14 12 152 12 * 28 58** ** 29 3 11 Total 327 genotypes from the parents spawned were sampled during the experiment (Tables I and 2). In 43 cases (4.9%). veligers could not be assigned to either population and were rejected. In 57 cases. no visible bands appeared (possibly indicating that only a shell, and not a live veliger, had been sampled). The proportions of progeny sampled from each female parent were consistent with the initial number of eggs spawned, and a comparison of the number of progeny inheriting each of the two alleles from each parent showed no selective mortality at this locus. A x^ analysis of the replicates by depth and age was performed. TABLE 3. Number of veligers at each depth and nominal depth for each mesocosm and sampling day. Age/Depth (m) Replicate #1 GEO PAS 14 d 0 20 2.5 11 5 12 8 0 Nominal depth 2,03 19 d 0 6 2.5 11 5 20 8 6 Nominal depth 4.08 24 d 12 33 d 0 19 2.5 19 5 18 8 4 Nominal depth 2.83 41 d (night) 0 16 2.5 14 5 24 g 13 Nominal depth 3.87 42 d 0 12 2.5 7 5 13 8 6 Nominal depth 3.43 15 8 0 2.50 123 8 3 5 2.59 7 11 10 10 2.76 13 13 5 9 3.24 12 9 2 0 1.41 Replicate #2 GEO PAS 4 13 10 20 5.16 32 18 17 20 10 3.42 15 12 20 3 3.08 22 7 6 1 1.54 16 14 10 3 2.53 18 10 8 7 2 2.63 10 4 7 3 2.88 7 19 5 0 2.34 The majority of the analyses showed little difference between the replicates with the following exceptions; at age 19 d. depths 5 (p = 0.008) and 8 (p = 0.037) m. which were not significant when Bonferroni testing is applied. The number of veligers is tabulated by population and depth in Table 3. In all but the first replicate at 14 d and the second replicate on day 42 (seven of nine cases), the nominal depth was greater for GEO than for PAS veligers. and when the two replicates are averaged together, the nominal depth was greater on four of five sampling days, which was consistent with the results of a previous experiment (Fig. 1 ). Ax" analysis of the depth distribution of populations by replicate and age was performed. In six of nine cases (all except age 33 d and replicate #2 at age 41 d). the two populations differed significantly or were marginal (p < 0.10) (Table 4). If we consider all nine tests to- gether, there are extremely significant (p < 0.001) differences in the depth distribution of these two populations of scallop veligers. DISCUSSION Although this experiment provided less detail about the vertical distribution of the veligers. it was less time consuming than re- trieval of data from the video tapes, and it demonstrated that three mesocosms is a reasonable level of replication for conducting ex- periments in the tower tank. In particular, it shows that differences in the vertical distribution of veligers are due to differences in veliger responses to the same stimuli, and not to small differences in food, light, water chemistry, etc.. or the chaotic amplification of the starting populations in replicated mesocosms. This meso- cosm replication technique provides significantly greater flexibil- ity in experimental design for tower tank experiments. The microsatellite probe developed by Gjetvaj et al. (unpub. obs.) has proved useful in linking genetics and behavior and has the potential to be useful in developing breeding protocols for P. magellanicus in hatchery situations. The great genetic variability in microsatellite probes makes it easy to choose parents with dif- TABLE 4. Probability (p for x" 'est I that the vertical distribution of veligers from different populations is the same. Age (d) Replicate #1 Replicate #2 14 19 33 41 42 0.094 0.004 0.998 0.067 0.010 0.000 0.634 0.489 0.003 256 Manuel et al. feait genotypes for pedigree analysis. These techniques could be u -d to determine whether different genotypes or broodstocks have different morlaUty over time or to allow testing of a number of broodstocks against different rearing techniques at the same time. In our experiments, we have invariably found that growth within a mesocosm was much less variable than growth rates among mesocosms. In hatcheries, where variability among containers is difficult to control, having several lines maintained in the same container would make comparison of lines very much easier and statistically more robust. In this experiment, the nominal depth of GEO veligers was greater than that of PAS veligers in six of nine samples taken. This agrees with the results of a previous experiment (Manuel 1996) where veligers from the PAS population did not migrate as deeply as did veligers from the GEO population. Interpretation of the significance of such differences requires that we distinguish be- tween genetic and environmental factors and between population and individual variation. In earlier experiments (Manuel 1996), it is conceivable that individual variation in either the genetics or the condition of the parents could have produced differences in the vertical distribution of veligers. The fact that GEO veligers were again found deeper than PAS veligers in this experiment increases the probability that the differences in vertical distribution were the result of genetics at the population level in two ways: the results were repeated in different years (so there is less possibility that something like the microposition of broodstock at the aquaculture site, resulting in perhaps different lipid levels in individual eggs, caused the differences), and we obtained the results from different adults from the same populations. In addition to the above, we used adults conditioned in a different manner (artificial vs. natural) and at a different time of year. It is therefore highly probable that there are real differences in vertical migration behavior between these two populations that are genetically based, and that the dif- ferences rest at the population as well as at the individual level. Manuel (1996) has suggested that scallop veligers are able to gain horizontal transport by migrating in the region of the ther- mocline. Thermoclines separate large bodies of water that often move in different directions and at different speeds. Even if hor- izontal transport is not the ultimate cause, it must be the result of vertical migration through such boundaries (Hill 199 1). Similarly, in two regions where hydrography differs, the same vertical mi- gration will have different consequences in terms of horizontal movement. If horizontal transport affects the survival of scallop veligers, then it follows that there will be selection for different vertical migration behavior in areas with different hydrography. Passamaquoddy Bay is an estuary with strong tidal influences, whereas Georges Bank is an offshore bank with a large clockwise gyre. Genetic differences in vertical migration behavior are con- sistent with horizontal transport influencing the vertical migration of scallop veligers. We also confirmed another phenomenon: GEO larvae are slower than PAS larvae to develop to the veliger stage. Although quantifying the difference was beyond the scope of this experi- ment, difference in development rate at the same temperature rep- resents another difference in larval ecology among populations. There may be a reason for the shorter nonfeeding trochopore stage of PAS veligers. The bay (PAS) population may be at greater risk of washout than the offshore (GEO) population. If veligers are able to affect horizontal transport by migrating vertically, but tro- chophores are not, then reducing the time speni in the trochophore stage would reduce the risk of washout. The differences in the development time of trochophores and the vertical migration behavior of veligers noted here warrant fur- ther investigation, both for the purpose of understanding the ecol- ogy of wild stocks and for choosing populations for aquaculture purposes. Aquaculturalists should at least be aware of differences in vertical migration behavior among different populations of po- tential aquaculture species. The movement of animals from one area to another where a local stock already exists may result in crossbred individuals that are not fit enough in the new system to sustain a breeding population. The seriousness of this type of scenario depends on three (at present) unknown factors: (1) the degree of difference among behaviors in the populations, (2) the heritability of those behaviors, and (3) the amount of mortality experienced as a result of "inappropriate"" behavior. On the other hand, the knowledge that such differences exist might be an ad- vantage. Chosing animals with appropriate veliger behavior (per- haps by using local parents) may greatly improve the success of stock enhancement programs. It is also possible that behaviors vary in their effect on survival or growth under the controlled conditions in the hatchery. If that were so, then simply culturing veligers for several generations may alter behavior, and strong selective pressure could severely increase inbreeding by removing all but the progeny that have inherited a particular chromosome from a particular individual. Strong selection for rare behavior within a limited gene pool could result in rapid inbreeding depression and consequent failure to thrive, negating attempts to improve broodstock by selection. Balch, N., C. M. Boyd & M. Mullin. 1978. Large-scale tower tank sys- tems. Rapp. P.-V. Reim. Cons. Int. E.xplor. Mer. 173:13-21. Demers, A. Y., Y. Lagadeuc, J. J. Dodson & R. Lemieux. 1993. Immu- notluorescence identification of early life-history stages of scallops (Pectinidae). Mar. Ecol. Prog. Ser. 97:83-89. Gallager. S. M., J. L. Manuel. D. A. Manning & R. K. O'Dor. 1996. Ontogenetic changes in the vertical distribution of scallop larvae Pin- copecten magellanicus in 9 m deep mesocosms as a function of light, food, and temperature. Mar. Biol. 124(4):679-692. Gjetvaj, B., M. Ball. S. Burbridge, C. J. Bird, E. Kenchington & E. Zouros. In prep. Characterization of microsatellite DNA variation in a natural population of the scallop Placopecten magellanicus. Harding. G. C, W. P. Vass, B. T. Hargrave & S. Pearre, Jr. 1986. Diel vertical movements and feeding activity ofzooplankton in St. George's Bay. N.S., using net lows and a newly developed passive trap. Can. J. Fish. Aquat. Sci. 43:952-967. LITERATURE CITED Hill, A. E. 1991. Vertical migration in tidal currents. Mar. Ecol. Prog. Ser. 75:39-54. Hurley, G. V., M. J. Tremblay & C. Couturier. 1987. Age estimation of sea scallop larvae (Placopecten magellanicus) from daily growth lines on shells. J. Northwest Atl. Fish. Sci. 7:123-129. Kaartvedt, S., D. L. Aksnes & J, K. Egge. 1987. Effect of light on the vertical distribution of Pecten maximus larvae. Mar. Ecol. Prog. Ser. 40:195-197. Manuel, J. L. 1996. Population and temporal variations in the vertical migrations of scallop (Placopecten magellanicus) veligers. Ph.D. The- sis. Dalhousie University. Halifax. Nova Scotia, Canada, 370 pp. Mam, K., A. Obara, K. Kikuchi & H. Okesaku. 1972. Smdies on the ecology of the scallop. Patinopecten yessoensis (Jay) 3. On the diurnal vertical migration of scallop larvae. Sci. Rep. Hokkaido Fish. E.xpl. Sin. 27:33-53. Patwary, M. U., E. L. Kenchington, C. J. Bird & E. Zouros. 1994. The Veligers of p. Magellanicvs 257 use of random amplified polymorphic DNA markers in genetic studies of the sea scallop Placopectcn magellcinicus (Gmelin. 1791 ). J. Shell- fish Res. 130:547-553^ Pearce. C M.. S M. Gallager, J. M. Manuel. D. A. Manning & R. K. O'Dor. 1996. Settlement of larvae of the giant scallop {Placopecten magellanicus) in 9-m deep mesocosms as a function of temperature stratification, depth, food, and substratum. Mar. Biol. I24(4):693- 706. Raby, D.. Y Lagadeuc. J J, Dodson & M Mingelbier. 1994. Relation- ship between feeding and vertical distribution of bivalve larvae in stratified and mixed waters. Mar. Ecol. Prog. Ser. 103:275-284. Scrope-Howe. S. & D. A. Jones. 1986, The vertical distribution of zoo- plankton in the western Irish Sea. Eslar. Coast. Shelf. Sci. 22:785- 802. Silva-Serra. M, A. & R, K O'Dor. 1994. Early life history traits of sea scallops. Placopecten magellanicus. from the Georges Bank popula- tion: vertical distribution of larvae. In: Proceedings of the Ninth Inter- national Pectinid Workshop. Nanaimo. B.C., Canada. April 22-27, 1993. 1:67-75. Sokal, R. R. & J. Rohlf. 1981. Biometry: the Principles and Practice of Statistics in Biological Research. W. H. Freeman, San Francisco. 859 pp. Tremblay, M. J. & M. Sinclair. 1990. Diel vertical migration of seal scallop larvae in a shallow embayment. Mar. Ecol. Prog. Ser. 67:19- 25. Journal of Shellfish Research. Vol. 15. No, 2. 259-264, 19%. THE USE OF LIPID EMULSIONS AS CARRIERS FOR ESSENTIAL FATTY ACIDS IN BIVALVES: A TEST CASE WITH JUVENILE PLACOPECTEN MAGELLANICUS PETER COUTTEAU,' JOHN D. CASTELL,^ ROBERT G. ACKMAN,' AND PATRICK SORGELOOS' ^Laboratory of Aquaculliire & Anemia Reference Center University of Ghent Rozier 44. B-9000 Gent. Belgium ^Department of Fisheries and Oceans Halifax Fisheries Research Laboratory P.O. Bo.x 550. Halifax. Nova Scotia. B3J 2S7 Canada ^Canadian Institute of Fisheries Technology Technical University of Nova Scotia P.O. Bo.x 1000. Halifax. Nova Scotia. B3J 2X4 Canada ABSTRACT Although information on bivalve nutrition is still very scarce, several studies have demonstrated the importance of lipids, in particular tnglycendes, as a source of energy and essential fatty acids in the early life stages. Expenmental diets used so far to study bivalve nutrition either heavily pollute the water or are too complex to prepare in a hatchery. The potential use of lipid emulsions as off-the-shelf supplements was evaluated through the analytical verification of the ingestion and incorporation of n-3 highly unsaturated fatty acids (HUFA) by the juvenile sea scallop Placopeclen magellanicus fed lipid emulsions of different fatty acid composition as a supplement to Isochrysis sp. (clone T-Isol. The average lipid content in the scallops fed the lipid supplements was 20% higher compared with that in the control fed algae only (3.29 ± 0.16 versus 2.75% of dry weight, respectively). Changes in the fatty acid composition, in particular of n-3 HUFA, were demonstrated in total lipids, polar lipids, and triglycerides of juvenile sea scallops supplemented with lipid emulsions on the basis of ethyl ester concentrates of n-3 HUFA and were dependent on the level and proportion of 20:5n-3 and 22:6n-3 present in the emulsion. The effective incorporation of essential fatty acids from lipid emulsions indicated that the supplementation of lipid emulsions to live algae may improve and standardize the dietary supply of lipids and fatty acids in hatchery production of bivalves. KEY WORDS: Bivalve, lipid, fatty acid, algal supplement, Placopecten magellanicus INTRODUCTION Rearing bivalves in commercial systems has so far relied on the production and use of selected species of unicellular marine algae. Despite the growing knowledge of the effects of environmental conditions, disease, and genetic background, the success of com- mercial hatchery cultures remains highly unpredictable. Mortali- ties are often attributed to deficiencies in certain nutritionally im- portant components. Although information on bivalve nutrition is still very scarce, several studies have demonstrated the importance of lipid quantity and quality, particularly triglycerides, in early life stages as a source of energy and essential fatty acids {Helm et al. 1973, Holland and Spencer, 1973, Waldock and Nascimento 1979, Gallager and Mann 1986, Gallager et al. 1986). A require- ment for the n-3 highly unsaturated fatty acids (HUFA), eicosa- pentaenoic acid (EPA; 20:5n-3), and docosahexaenoic acid (DHA; 22:6n-3), has been demonstrated for juvenile oysters (Langdon and Waldock 1981 ), and recent studies evaluating changes in fatty acid composition during larval development appear to confirm this for larval bivalves (Helm et al. 1991. Marty et al. 1992). The importance of lipids for larval development has encour- aged the development of artificial diets that provide specific lipid supplements during broodstock conditioning and larval rearing. Experimental diets used so far in bivalve nutrition studies, such as mixed diets (Trider and Castell 1980), liposomes (Parker and Se- livonchick 1986) and microcapsules (Langdon and Waldock 1981), either heavily pollute the water or are too complex to pre- pare on a regular basis in a hatchery. Lipid microspheres that are easily prepared by sonication of an oil mixture with lecithin and vitamin E have been proposed as a nutritional supplement for oyster conditioning (Robinson 1992a, Robinson 1992b. Heras et al. 1994). Self-emulsifying concentrates of marine oils, which are widely used in fish hatcheries to enrich filter-feeding prey organ- isms like Artemia and rotifers with n-3 HUFA (Sorgeloos and Leger 1992). are off-the-shelf lipid supplements that may also be acceptable for bivalves. Previous work has demonstrated the po- tential use of these lipid emulsions as a supplement for the larval bivalves Mercenaria mercenaria and Ostrea ediilis (Coutteau et al. 1994a). This study aimed at the analytical verification of the ingestion and incorporation of the fatty acids supplied through lipid emulsions of different n-3 HUFA content by juvenile Pla- copecten magellanicus . MATERIALS AND METHODS Juvenile sea scallops, P . magellanicus. were supplied by Fish- eries Resource Development Ltd. (Sandy Cove, Nova Scotia) and Andre Mallet (Mallet Associates, Halifax, Nova Scotia). The seed originated from stocks kept in a field nursery. During a period of 2 d, the animals were acclimated gradually to the experimental temperature (14— 15°C) and fed l.sochrisis sp. (clone T-Iso). Ini- tially, 3.5 g of scallops of approximately the same size (average live weight. 32.7 mg) were stocked per culture unit. The latter consisted of a small lantern net suspended in a bucket containing 25 1 of filtered (using cartridge filters with pore sizes of 5 and 1 |j.m) seawater that was renewed daily. Each bucket was aerated with an airstone to prevent the food from settling. After 17 d of feeding on the experimental diets, scallops were starved overnight 259 260 COUTTEAU ET AL. in iltered seawater. rinsed with distilled water, and stored at i2°C for biochemical analysis. Isochrysis sp. (clone T-Iso), which was selected as the algal control diet, was grown in lO-I batch culture with F/2 (Guillard) medium and harvested in the exponential phase. The scallops were fed an initial weight-specific daily ration of 0.88% (dry algae per initial wet seed biomass), which was administered over two feed- ings per day. The latter ration maintained algal concentration in the cultures above 15-20 cells |xP'. Rations were adjusted and the biomass was restocked after 6 d of culture in order to feed approximately constant weight-specific daily rations throughout the experiment (Urban et al. 1983). Algal rations were based on a dry weight of 13.8 ± 0.6 pg celP ' (mean and standard deviation from analysis of four cultures) for Isochrysis sp. (clone T-Iso), determined according to the method described by Coutteau et al. (1994b). Lipid emulsions were added simultaneously with the al- gae to give 0.2% lipid supplementation per initial wet scallop biomass. The experimental lipid emulsions (prepared by INVE Aquaculture N.V.-S.A., Belgium) contained, on a wet weight basis, 50% lipid, liposoluble vitamins (0.013% vitamin D,, 0.32% vitamin E, 0.08% ascorbyl palmitate, 0.18% vitamin A), emulsi- fiers, preservatives, antioxidants, and water. Lipid consisted of either coconut oil (EmO, control emulsion lacking HUFA) or ethyl ester concentrates of marine oils (Em50E and Em50D, approxi- mately 50% 2 n-3HUFA, primarily EPA and DHA, respectively). Lipids were extracted from whole animals (40-50 per extrac- tion), algae (four independent samples in the course of the exper- iment), and emulsions with a mixture of chloroformimethanol (2;1 v/v) (Folch et al. 1957). Total lipid contents were determined gravimetrically after exhaustive removal of the solvent from the lipid extract. Lipid classes were separated by preparative thin- layer chromatography on silica gel plates with hexane;diethyl ether:acetic acid (85; 15: 1), and fatty acid methyl esters (FAME) were prepared from total lipids, total polar lipids, and triglycerides by transesterification with 7% BFj-methanolibenzene (1:1 v/v) (Napolitano and Ackman 1993). Separation of the FAME was carried out on a Perkin-Elmer Model 8420 GC equipped with a flame ionization detector (FID) and an OMEGAWAX-10 flexible fused silica capillary column (30 m x 0.32 mm inner diameter) (Napolitano and Ackman, 1993). Relative areas were converted to weight percent amounts of fatty acids by correcting for the FAME FID responses (Ackman and Eaton 1978). Quantitative data (mil- ligrams of fatty acid per gram of dry weight, mg/g DW) were obtained by adding 10% of 23:0 as internal standard before the transesterification of total lipids and by using the following equa- tion: mgrA ■ ig DWy arecifA mg2io TL arfoijo CF ■ 100 with TL, total lipid sample (%DW); L, amount of lipid used for FAME preparation (g); CF, 1.04, correction factor for the con- version of fatty acids in FAME. Subsamples of scallops were dried at 70°C for 36 h to obtain the dry matter and then heated to 450°C for 4 h to obtain the ash weight. Organic matter was calculated f'rom the difference between dry matter and ash. RESULTS Isochrysis sp. (clone T-Iso) exhibited an average n-3 HUFA content of 14.0% with a DH A/EPA ratio of 17.8. The emulsions EmSOE and EmSOD contained approximately 50% n-3 HUFA with a DH A/EPA ratio of, respectively, 0.7 and 5.8. The coconut-oil based emulsion EmO contained 90% saturated fatty acids of which nearly 55% was 12:0 (Table 1). Scallops that were starved for 17 d did not show any increase in wet weight and exhibited higher ash content and lower lipid content compared with the fed ones (Table 2). The average lipid content in the lipid-supplemented treatments was 20% higher com- pared with that in the control fed algae only (3.29 ± 0.16 versus 2.75% of dry weight, respectively), whereas dry weight and ash content for all fed treatments were similar (in the range of 44.9- 47.8% and 82.2-85.7%, respectively). The proportions of monoenoic fatty acids increased during star- vation, mainly because of an increase of 20:1 (Table 3). A de- crease of n-3 fatty acids could mainly be attributed to the decrease of l8:3n-3 and particularly l8:4n-3, which was abundant in the initial scallops. Starved scallops selectively retained n-3 HUFA with C > 20, in particular 22:6n-3, and n-6 HUFA, especially 20:4n-6, whereas the proportion of n-3 HUFA with 20 C, primar- TABLE \. Fatty acid composition (weight % of total fatty acids) of the lipid emulsions and Isochrysis sp. (clone T-Iso). Fatty acid EmO EmSOE EmSOD ISO* 12:0 54.5 — 0.2 0.1 TMTD — — — 0.3 14:0 19.3 0.3 0.9 22.8 16:0 13,2 3.1 16.2 8.0 16:ln-9 — L3 1.8 0.3 16:ln-7 — — — 3.7 18:0 3.0 4.1 1.1 0.2 l8:ln-9 7.8 11.2 16.4 9.4 18:ln-7 0.1 3.4 1.4 1.2 18:2n-6 1.9 1.2 2.4 6.4 18;3n-3 — 0.9 0.4 7.0 18:4n-3 — 2.4 0.6 13.4 20:ln-ll — — — 3.0 20:ln-9 0.1 3.4 0.3 — 20;ln-5 — — — — 20:2n-6 — 3.7 4.1 0.1 20:4n-6 — L8 0.3 0.2 20:4n-3 — 1.5 0.1 — 20;5n-3 — 25.9 6.4 0.7 21:5n-3 — L6 0.9 0.5 22:2NMID — — — 1.4 22:ln-ll 4- n-13 — 1.6 — — 22:5n-6 — 0.9 1.3 2.9 22:5n-3 — 4.0 2.9 0.7 22:6n-3 — 19.3 37.5 12.1 24;ln-9 — 1.0 2.5 — S saturated 90.0 8.9 18.3 32.9 S monoenoic 8.1 24.1 22.2 17.9 S polyenoic 1.9 66.8 59.3 47.5 S n-3HUFA** 0 52.8 47.9 14.0 DHA/EPA — 0.7 5.8 17.8 Minor components identified (<1%) and not included in the table are lso-15;0; Ant-15;0; 15:0; lso-16:0; Ant-16:0; 7-methyl-hexadecanoic acid; 16:ln-5; 16;2n-6; l6:2n-4; 16:3n-4; 16;3n-3; 16;4n-l; 17:0 -I- phytanic (3,7,1 1,14-telramethylhexadecanoic) acid; 18:2n-4; 18:3n-6; 18;3n-4; 18:4n-l; 20:0; 20:ln-7; 20:2NMID; 20:3n-6; 20:3n-4; 20:3n-3; 22:0; 22;ln- 9; 22:ln-7; 22:ln-5; 22:4n-6; 22:4n-3. * Average values from analysis of four cultures. ** s=20:3n-3. Lipid Emulsions for Bivalves 261 TABLE 2. Average individual wet weight (WW) and composition of juvenile P. magellanicus at the start of the experiment and after 17 d of starvation or feeding on various diets. Treatment Average WW (mg ind~') Dry Weight (% WW) Ash 7c DW) Total Lipid {% DW) 2.01 L23 2.75 3.46 3.14 3.27 Initial After 17 d of culture Starved ISO ISO + EmO ISO + Em50E ISO + Em50D 32.7 33.0 47.5 54.1 53.9 57.0 45.7 44.0 46.1 45.7 47.8 44.9 ND 88.5 82.9 85.7 82.2 82.7 DW. dry weight; ND, not determined. ily 20;5n-3, decreased. Compared with the initial fatty acid com- position, scallops fed Isochrysls showed an increased proportion of certain monoenoic fatty acids that were also abundant in the alga, i.e.. 16:ln-7 and 18:ln-9, whereas the relatively high 20: ln-11 content in the algae did not affect that of the scallops (Table 3). The change of the HUFA content in total lipids after the ad- aptation to the Isochnsis diet reflected the HUFA composition of the alga, i.e.. a predominance of 22:6n-3 and a strongly reduced content in 20:5n-3. The supplementation of the emulsion EmO, despite its high content in 12:0. resulted in the presence of only 1% of 12:0 in the total lipids (Table 3). The proportion of total n-3 HUFA decreased slightly as the result of the EmO supplementation. The effect of the supplementation of the Em50 emulsions on the fatty acid compo- sition of total lipids was mainly restricted to an increase of the dominant n-3 HUFA. either EPA (from 3.3 to 5.6% for EmSOE) or DHA (from 17.9 to 19.4% for EmSOD). As a result, the sup- plementation of lipids with a lower DHA/EPA ratio than that of the algae resulted in a decrease of the DHA/EPA ratio in the scallop lipids from 5.5 in the control diet to 3.0 and 4.7 for the emulsions EmSOE and Em50D. respectively (Fig. lA). In accordance with the increase of the total lipid content and the proportion of n-3 HUFA. the absolute concentration of EPA and DHA was consid- erably higher in the scallops receiving the Em50 emulsions (re- spectively, 1.13 and 3.44 mg/g DW for ISO + Em50E, and 0.91 and 4.29 mg/g DW for ISO -I- EmSOD) compared with the control fed algae only (respectively, 0.54 and 3.00 mg/g DW) (Fig. lA). The supplementation of the EmO emulsion resulted in an increase of approximately 15% in the concentration of EPA and DHA. The above changes in the fatty acid composition of the total lipids due to lipid supplementation were amplified in the triglyc- erides (Fig. IB) and attenuated in total polar lipids (Fig. IC). In this way, the DHA/EPA ratio in the various dietary treatments was in the range of 3.2-8.1 and 2.8-3.7 for triglycerides and total polar lipids, respectively. Equivalent values for the total n-3 HUFA content were 15.5-23.0% and 34.7-36.6%, respectively. DISCUSSION Research on bivalve nutrition, including the study of lipid re- quirements, has been seriously hampered by the lack of suitable experimental diets. In this regard, lipid vesicles may have several advantages over other synthetic microparticles, e.g., their near- neutral buoyancy, suitable size range for efficient filtration by bivalves, and composition of nontoxic, digestible materials, Parker and Selivonchick (1986) demonstrated that juvenile Cras- sostrea gigas were able to metabolize the phosphatidylcholine and cholesterol present in the lamellae of liposomes. The protection from leaching of entrapped compounds makes liposomes particu- larly interesting as carriers for the delivery of water-soluble nutri- ents to filter-feeding organisms, whereas lipid microspheres con- sisting of emulsified lipid droplets provide a maximal amount of lipid per particle and may constitute effective carriers for lipid- soluble nutrients. Robinson (1992a) prepared microspheres of a lipid mixture of menhaden oil, egg phosphatidylcholine, and a partially hydrogenated vegetable oil with polyvinyl alcohol as an emulsifier. Although the latter author demonstrated the potential use of the lipid microspheres either as a supplement to or as a substitute for algae in the brookstock conditioning of the Pacific oyster, a similar lipid content and fatty acid composition was reported for nonfed oysters and oysters fed only the lipid micro- spheres (Robinson 1992b). Recently, Heras et al. ( 1994) improved the formulation of the emulsion of Robinson (1992a) by adding vitamin E and replacing the menhaden oil with a concentrate of n-3 fatty acid ethyl esters, thereby increasing the proportion of n-3 HUFA in the lipid supplement from 12.6 to 44.3%. Although it has been demonstrated with fluorescent beads that lipid micro- spheres are ingested and disintegrated by adult O. edidis (Heras et al. 1994). the latter does not prove the effective assimilation of nutrients from the lipid supplement. This work showed that es- sential fatty acids supplied as an emulsion of ethyl esters are as- similated and incorporated into the triglycerides and the polar lip- ids of juvenile sea scallops fed the lipid emulsion as a supplement to live algae. Similar work with O. edidis confirmed this for larval stages (Coutteau et al. 1994a), which may imply that lipid emul- sions could be used to provide dietary lipids to the various life stages and species of bivalve molluscs. The similar growth of scallops receiving a dietary supplement based on coconut oil and those fed a supplement rich in n-3 HUFA indicated that the algal control diet in this study, consisting of the DHA-rich Isochnsis. may have satisfied the requirement for n-3 HUFA in juvenile P. magellanicus. Nevertheless, the supplemen- tary dietary lipid may have provided additional energy, which was at least partially stored as lipid reserves in the tissues. The very limited accumulation of 12:0 in the total lipids (1% 12:0 in the tissue versus 54% in the emulsion), despite the increase of the total lipid content in the scallops fed the emulsion EmO, may indicate a rapid oxidation of a large part of the lipid supplement based on 262 COUTTEAU ET AL. TABLE 3. Selected fatty acid composition of total lipids of juvenile P. magellanicus at the start of the experiment and after 17 d of starvation or feeding on various diets (weight percent of total fatty acids). Fatty Acid Initial Starved ISO ISO + EmO ISO + EmSOE ISO + EmSOD 12:0 0.2 0.2 0.1 1.1 0.1 0.1 TMTD 2.7 — — — — — 14:0 2.9 1.5 8.3 8.9 8.3 8.0 15:0 0.5 1.1 0.4 0.3 0.3 0.3 16:0 15.2 14.3 11.8 11.8 11.9 11.8 16:ln-7 1.4 1.1 4.8 5.2 4.6 4.8 16:ln-5 0.6 1.0 0.1 0.1 0.1 0.2 16:3n-3 0.3 0.7 0.5 0.3 0.3 1.4 16:4n-l 0.3 1.9 0.9 1.3 1.2 0.1 17:0* 0.6 0.9 0.2 — 0.3 0.2 18:0 3.7 6.6 1.8 2.0 2.1 1.8 18:ln-9 3.4 3.5 8.0 8.4 8.1 8.1 18:ln-7 2.5 2.7 4.1 4.1 4.1 4.0 18:2n-6 2.9 1.3 5.1 5.3 5.0 4.9 18.3n-3 3.5 0.7 5.7 5.8 5.6 5.4 18:4n-3 12.6 1.9 13.5 13.5 13.2 13.2 20:ln-ll 1.1 3.4 1.0 1.1 1.0 0.9 20:ln-9 1.1 1.8 0.9 0.9 1.0 1.0 20:ln-7 0.7 1.4 0.4 0.5 0.4 0.3 20:ln-5 0.4 1.2 0.6 0.6 0.5 0.6 20:2n-6 2.0 1.8 1.3 1.5 1.2 1.3 20:3n-3 1.3 0.9 0.8 0.8 0.7 0.7 20:4n-6 1.1 3.4 1.0 1.1 1.0 1.1 20:4n-3 0.8 0.7 0.4 0.4 0.4 0.4 20:5n-3 13.0 10.1 3.3 3.0 5.5 4.1 21:5n-3 1.1 1.3 1.0 0.8 0.9 0.8 22:5n-6 0.6 1.2 2.7 2.2 2 2 2.5 22:5n-3 0.6 0.9 0.6 0.3 0.8 0.6 22:6n-3 19.0 26.3 17.9 16.3 16.7 19.4 S saturated 23.7 25.1 23.1 24.1 22.9 22.3 S monoenoic 11.7 17.1 20.5 21.6 20.5 20.4 S polyenoic 61.5 57.2 56.2 54.0 56.3 57.1 S n-3HUFA** 35.8 40.3 23.8 21.6 25.0 26.0 DHA/EPA 1.5 2.6 5.5 5.4 3.0 4.7 Minor components identified (<1%) and not included in the table are lso-16:0; 7-iiielhyl-hexadecanoic acid; 16:ln-9; 16:2n-6(n-4?); 16:3n-4; 16:4n-3; 20:0; 20:2NM1D; 22:2NMID; 22:4n-6; 24:0. * Includes phytanic (3,7,1 1.1 4-telraniethylhexadecanoic) acid. ** &20:3n-3. coconut oil. Similarly, small proportions of 12:0 have been ob- served in the muscle and liver lipids of sunshine bass fed coconut oil as main dietary lipid (Neniatipour and Gatlin 1993). A low accumulation of medium chain triglycerides (MCT) observed in ayu fish fed MCT (mainly 8:0) as a supplement to pollack liver oil has been attributed to the ability of fish to use MCT as a direct source of energy (Nematipour et al. 1989). The use of the coconut oil as a caloric supplement may have had a sparing effect on the catabolism of essential fatty acids provided through the algae, as indicated by the 15% increase of the concentration of EPA and DHA in the scallops fed the emulsion EmO as a supplement (Fig. lA). The conservative nature of the fatty acid composition of the polar lipid fraction compared with the triglycerides is widely ac- cci' d for fish (Sargent et al. 1993) and confirmed for various stages of bivalves (Waldock and Nascimento 1979, Langdon and Waldock 1981, Delaunay et al. 1993, Napolitano and Ackman 1993). The conservation of 20:1 fatty acids, in particular 20; In- 1 1 , observed in the starved and Isochn'sis-fed scallops is in agree- ment with its possible structural role in the gill membranes (Na- politano and Ackman 1993). The selective retention of 22:6n-3 and 20:4n-6 in the starved scallops, contrary to the decrease of 20:5n-3, indicated the greater importance of the former fatty acids. Similar observations on the relative importance of 20;4n-6 and 22:6n-3 versus 20:5n-3 can be deduced from other bivalve studies (Helm et al. 1991, Delaunay et al. 1993), and particularly, DHA may play an important role during larval development and meta- morphosis in the scallop Peclen maximus (Marty et al. 1992, De- launay et al. 1993). For bivalve larvae, it has been suggested that triglycerides may play a double role, first, as a storage of large amounts of saturated and monoethylenic fatty acids for energy purposes and, second, as a temporary reservoir of PUFA that could be transferred to struc- tural polar lipids and specific metabolic pathways during meta- morphosis (Napolitano et al. 1988). This is in agreement with several studies showing the importance of the content as well as the fatty acid composition of lipids, particularly triglycerides, for the early life stages of bivalves (Helm et al. 1973, Holland and Lipid Emulsions for Bivalves 263 Total I 25 « 20 ^- 15 - Z 10 '<^y ^\ Y> "<<> "x^ B: Triglycerides 'o. % <^o \ \ \ \ \ \ 8 7 o 6 ra 5 4 3 2 1 0 EPA DHA (n-3)HUFA DHA/EPA C: Total polar lipids Figure 1. Effect of starvation and supplementation of various lipid emulsions to Isochrysis sp. (clone T-lso) on the content of EPA, DHA, and n-3 HUFA in total lipids [(A) mg fatty acid/g of dry weight] and triglycerides and total polar lipids [(B and C) weight percent of total fatty acids] of juvenile P. magellanicus. Spencer 1973, Gallager and Mann 1986, Gallager et al. 1986, Helm et al. 1991, Ulting and Doyou 1992). Lipid emulsions may be a convenient artificial diet to study lipid and fatty acid require- ments during critical stages of bivalve culture, i.e.. broodstock conditioning and larval rearing. Furthermore, the ability of ma- nipulating both the quantity and the fatty acid composition of bivalve lipids, in particular of triglycerides, through the supple- mentation of off-the-shelf lipid supplements holds promise for application in hatchery rearing. Hatchery operators currently rely on live algae to supply essential fatty acids, and the fatty acid composition of algae may strongly vary with culture conditions (Pohl and Zurheide 1979). In this study, EPA and DHA concen- trations increased by >I00 and >40%, respectively, in juveniles fed lipid emulsions as a supplement to Isochrysis compared with the milligram/gram dry weight values in the control fed algae only. In particular, this accumulation of essential fatty acids through specific diet supplements could be used to standardize the dietary supply of lipids and essential fatty acids in broodstock condition- ing and larval reanng. ACKNOWLEDGMENTS This study was supported by the Belgian National Fund for Scientific Research (FKFO-project G. 2043. 90 and Postdoctoral Fellowship to Peter Coutteau), INVE Aquaculture N.V.-S.A., Belgium, and EU project contract CI 1-CT9 1-0945. Special thanks are due to Ms. Anne Timmins (TUNS, Canada) for her technical advice on the fatty acid analysis. LITERATURE CITED Ackman, R. G. & C. A. Eaton. 1978. Contemporary applications of open tubular gas liquid chromatography in analyses of methyl ester of longer chain fatty acids. Fette Seifen Anstrichmittel 80:21-37. Coutteau, P., M. Caers, A. Mallet, W. Moore, J. Manzi & P. Sorgeloos. 1994a. Effect of lipid supplementation on growth, survival and fatty acid composition of bivalve larvae iOslrea eduHs L. and Mercenaria mercenaria L.). pp. 213-218. In: P. Kestemont, J. Muir, F. Sevilla and P. Williot (Eds). Measures for Success. Proceedings of Bordeaux Aquaculture '94, March 23-25. 1994, France. CEMAGREF. Coutteau. P.. K. Cure & P. Sorgeloos. 1994b. Effect of algal ration on feeding and growth of juvenile Manila clam Tapes philippinarum. J. Shellfish Res. 1347-55. Delaunay, F., Y. Marty, J Moal & J. F Samain. 1993. The efrect of monospecific algal diets on growth and fatty acid composition of Pecien maximus (L.) larvae. J. Exp. Mar. Biol. Ecol. 173;163-179. Folch, J. M.. M. Lee & G. N. S. Stanley. 1957. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 29497-509. 264 COUTTEAU ET AL. Gaiiager, S. M. & R. Mann. 1986. Growth and survival of larvae of Mercenaria mercenaria (L.) and Crassostrea virginica (Gmelin) rela- tive to broodstock conditioning and lipid content of eggs. Aquacullure 56:105-121. Gallager, S. M., R. Mann & G. C. Sasaki. 1986. Lipid as an index of growth and viability in three species of bivalve larvae. Aquaciilriire 56:81-103. Helm, M. M., D. L. Holland & R. R Stephenson. 1973. The effect of supplementary algal feeding of a hatchery breeding stock of Ostrea edulis L. on larval vigour. J. Mar. Biol. Assoc. U.K. 53:673-684. Helm, M. M., D. L. Holland, S. D. Utting & J. East. 1991. Fatty acid composition of early non-feeding larvae of the European flat oyster, Oslrea edulis. J. Mar. Biol. Assoc. U.K. 71:691-705. Heras, H., J. Kean-Howie & R. G. Ackman. 1994. The potential use of lipid microspheres as nutritional supplements for adult Ostrea edulis. Aquacullure 123:309-322. Holland, D. L. & B. E. Spencer. 1973. Biochemical changes in fed and starved oysters. Ostrea edulis L. during larval development, metamor- phosis and early spat growth. J. Mar. Biol. Assoc. U.K. 53:287-298. Langdon, C. J. & M. J. Waldock. 1981. The effect of algal and artificial diets on the growth and fatty acid composition of Crassostrea gigas spat. J. Mar. Biol. Assoc. U.K. 61:431-448. Marty, Y., F. Delaunay, J. Moal & J. F. Samain. 1992. Changes in the fatty acid composition of Pecten maxirtius (L.) during larval develop- ment. J. Exp. Mar. Biol. Ecol. 163:221-234. Napolitano, G. E. & R. G. Ackman. 1993. Fatty acid dynamics in sea scallops Placopecten magellanicus (Gmelin, 1791) from Georges bank. Nova Scotia. J. Shellfish Res. 12:267-277. Napolitano, G. E., W. M. N, Ratnayake & R. G. Ackman. 1988. Fatty acid components of larval Ostrea edulis (L.): importance of triacyl- glycerols as a fatty acid reserve. Comp. Biochem. Physiol. 908:875- 883. Nematipour, G. R., H. Nishino & H. Nakagawa. 1989. Availability of medium chain tnglycerides as feed supplementation in Ayu Plecoglos- sus altivelis (Pisces), pp. 233-244. In: Proceedings of the Third Inter- national Symposium on Feeding and Nutntion in Fish. Aug. 28-Sept. 1, 1989, Toba, Japan. Nematipour. G. R. & D. M. Gatlin. 1993. Effects of different kind of dietary lipid on growth and fatty acid composition of juvenile sunshine bass, Morone chrysops x M. scLxatilis. Aquaculture 114:141-154. Parker. R. S. & D. P. Selivonchick. 1986. Uptake and metabolism of lipid vesicles from seawater by juvenile Pacific oysters {Crassostrea gigas). Aquacullure 53:215-228. Pohl. P. & F. Zurheide. 1979. Fatty acids and lipids of marine algae and the control of their biosynthesis by environmental factors, pp. 473- 523. In: H. A. Hoppe. T. Levring and Y. Tanaka (Eds.). Marine Algae in Pharmaceutical Science. Walter de Guyter, Berlin. Robinson, A. 1992a. Dietary supplements for reproductive conditioning of Crassostrea gigas Kumamoto (Thunberg). I. Effects of gonadal devel- opment, quality of ova and larvae through metamorphosis. J. Shellfish Res. 11:437^41. Robinson. A. 1992b. Dietary supplements for the reproductive condition- ing of Crassostrea gigas (Thunberg): II, Effects on glycogen, lipid and fatty acid content of broodstock oysters and eggs. J. Shellfish Res. 11:443-447. Sargent, J R., J G Bell, M. V. Bell, R. J Henderson & D. R Tocher. 1993. The metabolism of phospholipids and polyunsaturated fatty acids in fish. pp. 103-124. In: B. Lahlou and P. Viiiello (Eds). Aquacul- ture: Fundamental and Applied Research. Coastal and Estuarine Stud- ies. American Geophysical Union. Washington. D.C. Sorgeloos. P. & P. Leger. 1992. Improved larviculture outputs of manne fish, shrimp and prawn. J. World Aquacull. Soc. 23:186-217. Trider. D. J. & J. D. Castell. 1980. Effect of dietary lipids on growth, tissue composition and metabolism of the oyster (Crassostrea virgi- nica). J . Nutr. 110:1303-1309. Urban, E. R.. G. D. Pruder & Langdon, C. J. 1983. Effect of ration on growth and growth efficiency of juveniles of Crassostrea virginica (Gmelin), J Shellfish Res. 3:51-57. Utting. S. D. & J. Doyou. 1992. The increased utilization of egg lipid reserves following induction of tnploidy in the Manila clam {Tapes philippinarum). Aquacullure 103:17-28. Waldock, M. J. & I. A. Nascimento. 1979. The tnacylglycerol composi- tion of Crassostrea gigas larvae fed on different algae diets. Mar. Biol. Lett. 1:77-86. Journal of Shellfish Research. Vol. 15. No. 2, 265-270. 19%. ISOLATION AND CHARACTERIZATION OF A cDNA ENCODING AN ACTIN GENE FROM SEA SCALLOP (PLACOPECTEN MAGELLANIC US)* MOHSIN U. PATWARY,' MICHAEL REITH,' AND ELLEN L. KENCHINGTON^ t ^ Institute for Marine Biosciences National Research Council of Canada 1411 Oxford Street Halifax. Nova Scotia, Canada B3H 3Z1 ^Science Branch Invertebrate Fisheries Division Department of Fisheries and Oceans P.O. Box 550. Halifax. Nova Scotia. Canada B3J 2S7 ABSTRACT Two full-length complementary DNAs (cDNA) were isolated from a sea scallop adductor muscle-specific cDNA library and sequenced completely Both clones encode the same open reading frame of 376 amino acid residues. The amino acid sequence is highly homologous to other invertebrate actin sequences, and as in other invertebrates, the N-terminal sequences are much more similar to the nonmuscle actin of higher vertebrates than to their muscle actin. Results suggest that this is the primary actin gene expressed in adductor muscle. The size of the actin gene family in this species is approximately 12-15, determined through Southern hybndization. An actin gene probe may also be useful as a genetic marker because it reveals polymorphisms in sea scallop in at least three loci. Strong signals were obtained when DNA digests from several shellfish and fish were probed with an actin coding region probe, indicating that this clone will be useful in genetic studies of other fish and shellfish species. KEY WORDS: Actin, cDNA. genetic marker, Placopecien INTRODUCTION Actins are highly conserved contractile proteins. They are present in eukaryotic cells and play an important role in many diverse cellular process. They are globular proteins that polymer- ize into filaments for most of their biological functions, such as muscle contraction, cellular motility, and cellular structure (Pol- lard and Cooper 1986). Actins are divided into two classes, muscle and nonmuscle actins and, in vertebrates, are further grouped into three types: alpha actins. which occur in muscle tissues, and beta and gamma actins. which constitute the cytoskeletal actins present in most cells. Muscle and nonmuscle actins of vertebrates can be distinguished by a set of characteristic differences in amino acids (Vandekerckhoe and Weber 1984). In invertebrates, actins also have both muscular and nonmuscular functions, but these two classes are not readily distinguished on the basis of amino acid sequence. Actins are usually encoded by multiple genes in multicellular eukaryotes, with the largest number occurring in plants (Hennes- sey et al. 1993, Shah et al. 1983). These multigene families en- code different isoforms that are very similar in their primary struc- ture, yet they fulfill diverse functions in different organisms. They show tissue-specific expression at different stages of development. Promoter regions of the muscle-specific actin gene have been char- acterized and fused with reporter genes, and the tissue-specific expression of reporter genes has been reported (Macias and Sastre 1990, Pollard 1990). Because actin genes are very conserved at the amino acid level, they have been used to study codon bias and phylogenetic relationships across large phylogenetic distances (He *The sequence data reported in this article will appear in Genbank database under the accession No U55046. tAuthor for correspondence. and Haymer 1993, Fang and Brandhorst 1994). The studies on the molecular genetics of actin function have been reviewed by Hen- nessey et al. (1993). Although actin genes have been studied in a variety of organisms, they have not been studied at the molecular level in bivalves. The sea scallop (Placopecien magellanuus) is an important commercial species that occurs in discrete beds along the coasts of the northern United States and Atlantic Canada (Black et al. 1993). Because of its commercial importance, the sea scallop has been the subject of several genetic studies (Patwary et al. 1994a, Patwary et al. 1994b, Pogson and Zouros 1994, Volckaert and Zouros 1989). As a part of our continued interest in this species, we have constructed an adductor muscle-specific complementary DNA (cDNA) library, primarily to develop cDNA-based markers for population and aquaculture genetic studies of this organism. We have isolated and characterized the first bivalve full-length actin cDNA from our library. We also studied the potential use of this cDNA as a probe in genetic studies of sea scallop. MATERIALS AND METHODS DNA Extraction Adductor muscle samples were ground to powder in liquid nitrogen, immediately mixed with 10 mL/g tissue chilomonas buffer (2% sarcosyl, 0.5 M NaEDTA, 100 mM NaCl. 20 mM Tris IpH 7.6]) to which proteinase K (50-100 (xg/niL) was added, and incubated overnight at 55°C. Samples were then stirred briefly and centrifuged in an lEC clinical centrifuge (International Equipment company. Needham Heights. MA) at 1.975 Xg for 10 min; the supernatants were then transferred to a new set of tubes, and the nucleic acids were precipitated by adding two-third volume of cold isopropanol and maintaining for 1 h or more on ice or at 265 266 Patwary et al. -2iJ"C. The samples were centrifuged at 3.920 xg for 15 min. ana nucleic acid pellets were dried in a speed vac concentrator (Savant Instruments, Inc., Farmingdaie, NY) and dissolved in TE buffer (10 mM Tris-HCl, 1 mM EDTA) for 30 min at 55°C. Samples were treated with RNAse A (20 (xg/mL) for 30 min at 37°C and extracted twice with an equal volume of phenol- chloroform ((phenol ;chloroform:isoamyl alcohol = 50:49:1], 0.1 M Tris [pH 7.5], 0.2% p-mercaptoethanoll and once with an equal volume of chloroform-isoamyl alcohol (24:1). The DNA in the aqueous phase was precipitated by the addition of one-third vol- ume of NH4OAC and 2.5 volume of cold 957( ethanol for at least I h at -20°C and centrifuged at 10,000 xg for 15 min. The pellets were washed in 70% ethanol, dried, and dissolved in TE buffer (10 mM Tris-HCl, 0.1 mM EDTA). Preparation of RNA Total RNA was prepared by use of the methods described by Sures and Crippa (1984) and Turpen and Griffith (1986) with modifications. Twenty-one grams of adductor muscle tissue from one male and one female sea scallop was ground together to a powder in liquid nitrogen with a mortar and pestle, mixed with 190 mL of extraction buffer (4.0 M guanidine thiocyanate, 25 mM sodium citrate, 2% sarkosyl, and 1% p-mercaptoethanol), and incubated at room temperature for an hour. An equal volume of chloroform:isoamyl alcohol (24:1) was added to the sample, mixed for 10 min, and centrifuged to separate the two phases. CsCl (0.2 g/mL) was added to the aqueous phase, layered over a CsCI cushion (5.7 M CsCI, 50 mM EDTA), and centrifuged at 184,000 Xg for 6 h in a 70 Ti rotor (Beckman Instruments, Inc., Palo Alto, CA). The pellets were air dried and dissolved in ME buffer [10 mM 3-(N-morpholine) propanesulfonic acid, 4.5 mM EDTA, pH 7.2 0.17f diethyl pyrocarbonate (DEPC)] containing 0.5% sodium dodecyl sulfate (SDS) and 5% phenol, and RNA was extracted with phenol-CHCl, and CHCl, and precipitated with ethanol at - 70°C for 30 min. The sample was centrifuged. and the RNA pellet was dried in a speed vac concentrator without heat on and dissolved in DEPC-treated water. Polyadenylated RNA was isolated from total RNA by the use of a Poly(A) Quick mRNA purification kit according to the sup- plied protocol (Stratagene. La Jolla. CA). A total of 20.0 |jLg of poly A* RNA were obtained from approximately 1.000 [j,g of total RNA. The cDNA was synthesized, size fractionated, and ligated into Uni-ZAP vector arms according to the protocol supplied with the ZAP-cDNA synthesis kit (Stratagene). A total of 2.5 |xg of mes- senger RNA yielded a library of 1.9 x 10^ pfu. Library Screening A few microlitres of the library were plated, and a total of 130 plaques were randomly cored and stored in SM medium (0.58% NaCl, 0.2%; MgSOj • 7H,0, 50 mM Tris-HCl (pH 7.5], 0.01% gelatin) at 4°C. The Bluescript phagemids from 12 plaques were excised in vivo from lambda-ZAP 11 with ExAssist helper phage. Phagemid DNA was extracted by the use of a Nucleobond AX kit ■ Macherey-Nagel GmbH & Co.. Duren. Germany). The clones containing inserts were partially sequenced in both directions with an ABI 373 sequencer (Applied Biosystems, Foster City, CA), and four were identified as actin clones through Blast P searches (Altschul et al. 1990). Additional actin clones were isolated through subsequent screening with an actin coding region probe. and the two largest clones were completely sequenced in both directions. Preparation of Probe A nonradioactive actin probe was prepared by labeling with alkali-labile (Dig-1 1-dUTP. Labeling was done either by the poly- merase chain reaction (PCR) or by the random labeling method. In PCR labeling, the 25-|jlL reaction contained 1 x Taq polymerase buffer: 1 .5 mM MgCL; 100 p.M each of dATP, dCTP. and dGTP; 86 |xM dTTP; 14 jxM dig-II-dUTP; 25 ng each of forward and reverse nested primer; 1 U of Taq DNA polymerase-; and 5 ng of recombinant plasmid DNA or 1 jxL of frozen and thawed recom- binant Uni-ZAP XR vector in SM medium. The reactions were performed in a GeneAmp 9600 PCR system (Perkin Elmer, Nor- walk. CT) and programmed as follows using the fastest ramp time: 1 cycle at 94°C for 4.5 min. followed by 29 cycles each of 30 s at 94°C. 30 s at 60°C. and 4 min at 72°C. and a final cycle of 30 s at 94°C. 30 s at 60°C. and 10 min at 72°C. Successful labeling is indicated by decreased mobility of the reaction product in an aga- rose gel compared with an unlabeled control. Random primed labeling of cDNAs was done with a Boehringer Mannheim (Boeh- ringer Mannheim. Laval. PQ. Canada) kit. To avoid common flanking M13 sequences from the probe, a pair of nested primers for each clone was synthesized and used for insert amplification. These amplified inserts were purified with a QIAEX Gel Extrac- tion kit (Qiagen) and used as a template for random labeling. The probe was cleaned either by using Nick columns (Pharmacia Bio- tech. Uppsala, Sweden) or by using the QIAquick Spin PCR Pu- rification Kit (Qiagen. Chatsworth. CA) and quantified by com- paring with known Dig-labeled DNA on a dot blot. Preparation of Genomic Blots DNA samples ( 10 |xg each) were digested overnight with 50 U of appropriate restriction enzymes. To aid in DNA digestion, sper- midine was added to a final concentration of 5 mM in the reaction. The digested DNAs together with digoxigenin-labeled molecular- weight marker III (Boehringer Mannheim) were fractionated in 0.8% agarose gels at 20 V/cm in Tris-acetate (0.04 M Tris-acetate, 0.001 M EDTA) buffer and were then transferred to positively charged nylon membranes (Boehringer Mannheim) with a Phar- macia vacuGene XL unit and following the manufacturer's proto- col no. 1. Hybridization Prehybridization was done in a hybridization oven at 39°C for 4 h in a buffer containing 50%: deionized formamide. 5x sodium chloride, sodium citrate (SSC), 0.1%c N-lauroylsarcosine, 0.02% SDS, and 2% blocking reagent (Boehringer) and 100 (ji,g/mL boiled and chilled yeast RNA (Boehringer). Hybridization was done under the same conditions for 1 8 h in fresh buffer to which 10 ng/mL denatured probe was added. Blots were washed at room temperature twice in 2 x SSC-0.2% SDS for 10 min, once in 0.5 x SSC-0. 1% SDS for 15 min. and once in 0.2x SSC-0. 1%. SDS for 30 min. To detect Dig-labeled DNA by chemiluminescence with Lumigen PPD. the Boehringer Mannheim protocol was followed, except that the membrane was incubated in 1 .25% buffer 2 for 90 min, instead of 30 min in 1%> buffer 2, and anti-Dig-alkaline phosphatase was diluted 1:12,500 instead of 1:10,000. The mem- branes were sealed in polythene bags and exposed to X-ray film. 1 gagcacccacaccaatagtctttagactcaccttggaactaac_cccaacaaccaacaaacaaac ATG TGT GAC GAC 76 1 M C D D 4 77 GAG GTA GCA GCT TTA GTA GTA GAC AAT GGC TCC GGT ATG TGC AAG GCC GGG 1TC GCC GGA 13 6 5EVAALVVDNGSGMCKAGFAG 24 137 GAC GAT GCT CCA CGC GCT GTG VTC CCC TCC ATT GTT GGA AGG CCC CGT CAC CAG GGT GTC 196 25DDAPRAVFPSIVGRPRHQGV 44 197 ATG GTT GGT ATG GGT CAG AAA GAC AGC TAC GTA GGA GAT GAA GCT CAG AGC AAG AGA GGT 256 45MVGMGQKDSYVGDEAQSKRG 64 257 ATC CTC ACC CTC AAG TAC CCC ATT GAG CAC GGT ATC GTC ACA AAC TGG GAT GAT ATG GAG 316 65ILTLKYPIEHGIVTNWDDME 84 317 A-AG ATC TGG CAT CAC ACC TTC TAC AAC GAG CTC CGT GTC GCC CCT GAG GAG CAC CCC GTC 37 6 85KIWHHTFYNELRVAPEEHPV 104 377 CTC CTG ACA GAG GCT CCC CTC AAC CCC AAG GCC AAC AGG GAA AAG ATG ACC CAG ATC ATG 436 105LLTEAPLNPKANREKMTQIM 124 437 TTC GAG ACC TTC AAC GCC CCC GCT ATG TAC GTC GCC ATC CAG GCT GTC CTC TCC CT3 TAC 496 125FETFNAPAMYVAIQAVLSLY 144 497 GCT TCC GGT CGT ACC ACC GGT ATC GTC CTC G;>.C TCC GGA GAT GGT GTC ACC CAC ACC GTC 556 145 ASGRTTGIVLDSGDGVTHTV 164 557 CCC ATC TAT GAA GGT TAC GCT CTT CCC CAC GCC ATC CTC CGT CTC GAC TTG GCT GGC CGT 616 165PIYEGYALPHAILRLDLAGR 184 617 GAC TTG ACC GAT TAC CTC ATG AAG ATC CTC ACC GAG CGT GGT TAC TCA TTC ACC ACC ACC 67 6 185 DLTDYLMKILTERGYSFTTT 204 677 GCC GAG AGA GAA ATC GTC AGG GAC ATC AAG iGAG AAA CTC TGC TAT GTT GCC CTC GAC TTC 73 6 205 AEREIVRDIKEKLCYVALDF 224 737 GAG AAC GAG ATG GCC ACC GCC GCC TCA TCC TCA TCC CTC GAG AAG AGC TAC GAG CTT CCC 7 96 225 ENEMATAASSSSLEKSYELP 244 797 GAC GGT CAG GTC ATC ACC ATC GGA AAC GAG CGT TTC AGG TGT CCC GAA TCC CTC TTC CAG 856 245 D.GQVITIGNERFRCPESLFQ 264 857 CCA TCC TTC TTG GGT ATG GAA TCT GCC GGT ATC CAC GAG ACC ACA TAC AAC TCC ATC ATG 916 265 PSFLGMESAGIHETTYNSIM 284 917 AAG TGC GAC GTC GAC ATC CGT AAG GAT CTG TAC GCC A-AC ACT GTC CTG TCC GGA GGC ACC 97 6 285 KCDVDIRFCDLYANTVLSGGT 304 977 ACC ATG TTC CCA GGT ATT GCC GAT CGT ATG CAG AAG GPA ATC ACC GCC TTG GCT CCC AGC 1036 305TMFPGIADRMQKEITALAPS 324 1037 ACA ATG AAG ATC AAG ATC ATT GCP CCA CCA CJAG AGG AAA TAC TCC GTC TGG ATC GGT GGC 1096 325 TMKIKIIAPPERKYSVWIGG 344 1097 TCC ATC TTG GCT TCT CTG TCC ACC TTC CAA CAG ATG TGG ATC AGC AAA CAG GAA TAC GAT 1156 345 SILASLSTFQQMWISKQEYD 364 1157 GAG TCC GGC CCA TCC ATT GTC CAC AGG AAA TGC TTC TA.A atCatCcttcaaattatCaggacCtcaa 1223 365 ESGPSIVHRKCF' 377 1224 attaCttttacattttcggtactgtgaagaaCggactccacctcgtgctctattgaacaggaCaagctgacagacagaaC 1303 1304 ctcgtcttgcccgataaagtgcgatacctaagtgctcttaaaaagacaccagcgacatcccgtcagcagcctgggtaagg 1383 1384 caatgtttcgaaggttcctaccaaccaaagtacacggacagctatgcacactaaggacacctgccgtcctttcatctaat 1463 14 64 ataatttaacttaaataaatctacgcaaatctacagctcgatggaagttgatttcccaggacgggaagttacactttgta 1543 1544 caggaagttgagagtatgacgggacgtactgtgtacacccgcgtaactatggtgatgtaatgtatacagcactcacacac 1623 1624 Cgtaagaaaaactggcatcatttctcaaagaacatcgtgataacgcacccacgcctgcatacatgccgcatggatgtagg 1703 1704 Cccaccatccagagcaaactaaaaatgttaaagtctaaaaaaaaaaaaaaaaaaaa 1759 Figure I. Nucleotide sequence and deduced amino acid sequence of the sea scallop actin gene. The nucleotide residues are numbered from the 5' end of clone PmC-ll, and the amino acid residues are numbered from the first in-frame methionine. The nucleotide sequence shown was determined on both strands of the entire cDNA insert of /"mC-ll and PmC-4. The potential polvadenylation signal is underlined. The nested primers that were used to amplify this clone are identified by broken underlines. 268 Patwary et al. kb M a a' b c d e f 21.2 5.2 3.5 # 2.0 9 t • • • 0.6 Figure 2. Detection of actin genes in the sea scallop. Sea scallop ge- nomic DNA blot hybridized with an actin coding region probe. DNA from a single animal was digested with the following restriction en- zymes: lanes a and a', EroRl (no spermidine added in a'); lane b, £foRV; lane c, Hindlll; lane d, Ndel; lane e, Pstl; and lane f, Xbal. M is digoxigenin-labeled DNA molecular-weight marker III (Boehringer Mannheim). RESULTS AND DISCUSSION From 130 randomly chosen Uni-Zap XR recombinant lambda plaques. 12 inserts were partially sequenced. Four of these inserts were identified through BlastP searching as encoding actin. By the use of one of these inserts (PmC-l 1 ) as a probe, an additional 30 actin clones were identified from among the 118 remaining plaques. Actin cDNAs were found to be the most common (29%) in our library. The other cDNAs that occurred frequently were myosin (15%). arginine kinase (6%), and tropomyosin {4%). The hybridization of a probe prepared from the PmC-\l noncoding region to the actin clones indicated that they all represent the same gene. This observation was also supported by the complete se- quence homology in the 3' noncoding regions of several clones. These data indicate that, as in sea urchin (Lee et al. 1984). only a single actin gene is expressed in scallop adductor muscle. Complete sequencing of the two largest clones showed that they were 1,750 and 1.759 base pairs (bp) long. The shorter clone was missing 9 bp at the 5' end. The largest clone had 64-bp and 544-bp (excluding poly As) noncoding regions at the 5' and 3' ends, respectively. Both clones encoded the same open reading frame of 376 amino acid residues (Fig. 1). Actin sequences are highly conserved at the amino acid level in evolution. The sea scallop sequence differs by only 6 amino acids from the California sea hare (sp/P17304); by 7 amino acids from Anemia (sp/P18603); by 8 amino acids from Caenorhahdilis (Pir/ s27135); by 10 amino acids from Drosophila (sp/P10987) and silkworm (sp/P04829); by 1 1 amino acids from common carp (sp/P12714). chicken (gp/M10279). mouse (pir/A31900). and starfish (sp/P12716); and by 13 amino acids from sea urchin (sp/ P02573). Owing to this level of similarity, actin amino acid se- quences may not be useful to study phylogenetic relationship among invertebrates. The predicted N-terminal sequences are sim- ilar to those of other invertebrates. As in other invertebrates, the first five amino acids in sea scallop are methionine-cysteine and three acidic amino acids — a valine at position 1 1 , methionine at position 17. and a cysteine at position 18. As with other inverte- brates, the N-terminal sequences are much more similar to non- muscle actin of higher vertebrates than to their muscle actin (Rubenstein 1990). Southern blot analysis was done to estimate the number of genes encoding actin in the sea scallop. Ten micrograms of DNA from a single sea scallop was digested overnight with each of the six enzymes £(V)R1. EcoR\ . HmdXW. Ndel. Pstl. and Xbal. elec- trophoresed. and blotted. Hybridization of the blot with an actin coding region probe produced signals in all lanes (Fig. 2). The same signal pattern was obtained with DNA from a different an- kb 21.2 Mabcdefghii 5.2 2.0 ! 0.6- Figure 3. f//ncll-digested genomic DNA blot hybridized with an actin coding region probe. Each lane contains DNA from a dilTerenl animal. The animals are: lanes a and b from Yarmouth, Nova Scotia: lanes c, d, g, and h from Sable Island; lanes e and f from an unknown location: and lanes i and j from Annapolis Basin. Nova Scotia. M is a DNA marker as mentioned in the legend to Figure 2. Arrows indicate poly- morphic loci. The apparent polymorphic high-molecular-weight faint bands are possibly the results of incomplete digestion. p. MAGELLANICUS ACTIN cDNA 269 kb Mabcdefgh 21.2 ^ 5.2 w ♦« s 2.0- S • If 0.6 - • Figure 4. ///ncll-digested genomic DNAs from different species of mollusc and Fishes hybridized with actin probe. DNAs are from: lane a, P. magellaniciis (sea scallop): lane b, Chlamys hastata (Icelandic scallop): lane c, Mytilus edulis (blue mussel); lane d, Spisiila solidissima (surf clam); lane e, Homarus americanus (American lobster); lane f, Gadus morhiia (cod); lane g, Tilapia nilotica (tilapia); and lane h, Crassadoma gigantea (rock scallop). M is a DNA marker as mentioned in the legend to Figure 2. Note that signals in lane d are very weak because only a very small quantity of DNA was available to load. imal. The Southern blot data suggest the presence of approxi- mately 12 to 15 actin genes in the sea scallop. The sea scallop appears to have a slightly larger family size than other inverte- brates such as sea urchin. Artemia. and Drosophila. which have about 8. 8-10, and 6 genes, respectively (Singer and Berg 1991). To determine the potential use of actin as a genetic probe, genomic DNAs from 10 sea scallops, 8 from three different beds and 2 from an unknown location, were digested with HmcII, blot- ted, and hybridized with an actin cDNA coding region probe (Fig. 3). Although most of the actin-containing DNA fragments re- vealed by this blot are constant among individuals tested, there are at least three polymorphic loci at 4.3, 3.0, and 2.5 kilobases (kb). The 4.3-kb locus is highly heterozygous, with apparently five different alleles occurring at this site. The 3.0-kb locus is highly homozygous in these samples with only one (lane a) having a heterozygous allele. The 2.5-kb locus is equally homozygous and heterozygous. An additional polymorphic locus may occur at ap- proximately 6.5 kb, although analysis of additional individuals is necessary to confirm this. Although these results are too limited to make any genetic inference about sea scallop populations, they demonstrate that the actin probe may be useful for a variety of genetic studies. The sea scallop actin gene probe was hybridized with HincU- digested sea scallop, Icelandic scallop, rock scallop, blue mussel, surf clam, American lobster, cod, and tilapia adductor muscle on a genomic DNA blot (Fig. 4). In all cases, signals of good mten- sity were obtained. These results reflect the conserved nature of the actin gene and suggest that it can be used as a reference in gene expression studies and as a heterologous probe to isolate actin genes, as well as to conduct genetic studies on these organisms. The sea scallop actin cDNA characterized here appears to rep- resent the primary, and perhaps only, actin gene expressed in the adductor muscle. This cDNA will be a useful tool for isolating sea scallop cytoskeletal actin genes and cDNAs, as well as actins from other marine organisms. Interestingly, sea scallop actin genes also appear to be useful genetic markers, although it is not yet clear at what level (individual or population) they will be most useful. ACKNOWLEDGMENTS We gratefully acknowledge the NRC Institute for Marine Bio- sciences (1MB), Halifax, Nova Scotia, Canada, for providing fa- cilities and supplies to M.U.P. to conduct this work. This project was supported in part by funds from the Department of Fisheries and Oceans through a contract to the NRC Institute for Marine Biosciences. We thank Mr. D. Cook (Marine Gene Probe Labo- ratory, Department of Biology, Dalhousie University) for provid- ing cod and tilapia DNA samples. We also thank Ms. Carolyn J. Bird of 1MB for reviewing the manuscript and suggesting im- provements. Issued as NRCC No. 39703. LITERATURE CITED Altschul.S. F.,W.Gish,W. Miller. E. W. Myers & D. J. Lipman, 1990. Basic local alignment search tool. J. Mol. Biol. 215:403-410. Blactc. G. A. P., R. K. Mohn, G, Robert & M. J. Tremblay. 1993. Atlas of the biology and distnbution of the sea scallop Placopeclen magel- laniciis and Iceland scallop Chlanns islandica in the Northwest Atlan- tic. Can. Tech. Rep. Fish. Aqiial. Sci. No. 1915:1-34. Fang, H. & B. P. Brandhorst. 1994. Evolution of actin gene families of sea urchins. J. Mol. Evol. 39:347-356. He, M. & S. Haymer. 1995. Codon bias in actin niultigene families and effects on the reconstruction of phylogenetic relationships. J. Mol. Evol. 41:141-149. Hennessey, E. S., D. R. Drummond & J. C. Sparrow. 1993. Molecular genetics of actin function. Biochem. J. 282:657-671. Lee, J. J., R. J. Shott. S. J. Rose, T. L. Thomas, R. J. Britten & EH. Davidson. 1984. Sea urchin actin gene subtypes. Gene number, link- age and evolution. J. Mol. Biol. 172:149-176. Macias, M.-T & L. Sastre. 1990. Molecular cloning and expression of four actin isoforms during Anemia development. Nucleic Acids Res. 18; 5219-5225- Patwary. M. U., R. M. Ball, C. J. Bird. B. Gjeuaj. S. Sperker. E. Kench- ington & E. Zouros. 1994a. Genetic markers in the sea scallop and their application to aquaculture. Bull. Aquacul. Assoc. Canada 2:18-20. Patwary. M. U., E. L. Kenchington, C. J. Bird & E. Zouros. 1994. The use of random amplified polymorphic DNA markers in genetic studies of the sea scallop Placopeclen magellanicus (GMELIN, 1791). J. Shellfish Res. 13:547-553. 270 Patwary et al. Pogon. G. H. & E. Zouros. 1994. Allozyme and RFLP heterozygosities as correlates of growth rate in the scallop Placopecien magellanicus: a test of associative overdominance hypothesis. Genetics 137:221-231. Pollard, T. D. 1990. Actin. Curr. Opin. Cell Biol. 2:33-40. Pollard. T. D. & J. A. Cooper. 1986. Actin and actin-binding proteins. A critical evaluation of mechanisms and functions. Aimu. Rev. Biochem. 55:987-1035. Rubenstein, P. A. 1990. The functional importance of multiple actin iso- forms. BioEssays 12:309-315. Shah. D. M.. R. C. Hightower & R. B. Meagher. 1983. Genes encoding actins in higher plants are highly conserved but coding sequences are not. J. Mol. Appl. Genet. 2:111-126. Singer. M. & P. Berg. 1991. Genes and Genomes— a Changing Perspec- tive. University Science Books, Mill Valley, California. 929 pp. Sures. 1. & M. Cnppa. 1984. Xenopsin: the neurotensin-like octapeptide from Xenopus skin at the carboxyl terminus of its precursor. Proc. Natl. Acad. Sci. U.S.A. 81:380-384. Turpen. T. H. & O. M. Gnffith. 1986. Rapid isolation of RNA by a guanidinium thiocyanate/cesium chlonde gradient method. BioTech- niqites 4:1 1-13. Vandekerckhoe , J. & K. Weber. 1984. Chordate muscle actins differ distinctly from invertebrate muscle actins. J. Mol. Biol. 179:391- 413. Volckaert, F. & E. Zouros. 1989. Allozyme and physiological variation in the scallop Placopecten magellanicus and a general model for the effect of heterozygosity on fitness in marine molluscs. Mar. Biol. 103:51-61. Journal of Shellfish Research. Vol. 15. No. 2, 271-283. 1996. FOOD-LIMITED GROWTH AND CONDITION INDEX IN THE EASTERN OYSTER, CRASSOSTREA VIRGINICA (GMELIN 1791), AND THE BAY SCALLOP, ARGOPECTEN IRRADIANS IRRADIANS (LAMARCK I8I9) ROBERT B. RHEAULT' AND MICHAEL A. RICE^ ^Moonstone Oysters 1121 Mooresfield Rd. Wakefield. Rhode Island 02879 ^Department of Fisheries. Animal and Veterinary Science University of Rhode Island Kingston, Rhode Island 02881 ABSTRACT The growth response of the eastern oyster. Crassoslrea virginica. and the bay scallop. Argopecten irraJians irradians. to varying degrees of food limitation was evaluated. Under conditions of low current speed, dense assemblages of shellfish can rapidly deplete ambient food concentrations, resulting in measurable effects on growth and condition index. A flume study demonstrated significant growth and condition index responses to resource competition after reductions as small as ll'Jc in relatively high ambient food concentrations (=4.6 (ig/l chlorophyll). Growth rates and condition index are linearly correlated with the average chlorophyll ration consumed. A field study demonstrated similar growth responses when the shellfish were cultured over a range of densities in a commercial aquaculture setting. By comparing the growth and condition index responses in the two expenmenls. we infer the degree of resource depletion occurring in the field from the correlations constructed in the flume study. Although physiological responses to food limitation will necessarily be site specific to varying combinations of temperature, current speed, and food concentration or quality, this work provides a unique opportunity to compare the growth response of oysters and scallops under a wide range of food availability in both laboratory and commercial aquaculture settings Doubling the stocking density from 2.5 to 5.0 kg of oysters per bag resulted in a 20% decrease in both the condition index and the growth rate (percent increase in weight). These observations may assist commercial growers determine optimal slocking density for their aquaculture grow-out systems. Natural food availability in Point Judith Pond, a classic salt wedge estuary, is highly variable on a daily basis and is related to the tidal exchange. The variation in food concentration superimposed on the tidal current oscillation leads to massive changes in food flux and the degree of local resource competition. Scallop and oyster clearance rates (milliliters per minute) were constant over a wide range of chlorophyll concentrations, suggesting that these species will filter natural seston at a near-constant rate despite fourfold tidal variations in food concentrations. Scallop clearance rates were reduced when chlorophyll concentrations were depleted to below 12% of the natural levels, suggesting a threshold feeding response. KEY WORDS: Bivalve, growth, condition index, culture, seston flux, tlume INTRODUCTION 1981. Malinowski and Siddall 1989. Newell 1990). The world- wide economic importance of shellfish aquaculture is increasing In the marine environment, two basic factors influence the because many wild-harvest fisheries have peaked or are in decline availability of food to benthic suspension feeders. The primary (FAO 1992). Critical to the success of shellfish aquaculture oper- factor is the concentration of phytoplankton and particulate or- ations is an understanding of the effects of stocking density on ganic matter in the water column. For sparse populations, or where growth rate (Incze et al. 1981. Newell et al. 1989). Optimal stock- populations are exposed to strong currents, the concentration of ing density in commercial aquaculture is determined by site- food is the only factor controlling food availability. However, specific physical factors (horizontal seston flux), biological factors when populations of suspension feeders occur at great densities, or (species-specific filtration rate), and economic factors such as gear when currents are insufficient to replenish the food, local resource and labor costs. Lower stocking densities will almost always result competition can deplete the concentration of locally available food in less competition for food and faster growth rates, but these (Dame et al. 1984. Frechette et al. 1989. Peterson and Black increases come at the expense of more investment in gear, more 1991). Food availability for these populations is determined by a labor to maintain it. and larger lease requirements, product of both the food concentration and the current speed, or This article describes the results of experiments designed to horizontal seston flux (Muschenheim 1987. Grizzle and Lutz elucidate the effects of stocking density on food limitation in a 1989, Muschenheim and Newell 1992). modified rack-and-bag shellfish aquaculture operation (Rheault There is evidence that local food depletion can result in food- and Rice 1995). We used tightly controlled flume studies to quan- limited growth in natural bivalve assemblages (Wildish and Krist- tify the effects of downstream food depletion on growth and con- manson 1985. Frechette and Bourget 1985, Rice et al. 1989); dition index in eastern oysters Crassoslrea virginica (Gmelin however, the subject has been the focus of considerable debate 1791) and northern bay scallops Argopecten irradians irradians (Reiswig 1971 . Powell et al. 1987). Although local food depletion (Lamarck 1819). The results of the flume study were then com- in natural assemblages may be the exception rather than the rule. pared with growth and condition index data from a concurrent field there is little doubt that food depletion occurs at population den- experiment that used commercial aquaculture techniques over a sities common to commercial shellfish aquaculture. Several stud- range of initial stocking densities. ies have described intensive aquaculture systems that suffer from Other researchers have used flumes to examine food depletion food-limited growth (Duggan 1973. Mason 1976, Rhodes et al. in filter-feeding bivalves (Asmus and Asmus 1991. Butman et al. 271 272 Rheault and Rice 1994 and references therein). Many of these experiments were conducted with cultured phytoplankton (Kirby-Smith and Barber 1974, Eckman et al. 1989, Cahalan et al. 1989) or at population densities insufficient to cause food depletion (Grizzle et al. 1992. Wildish et al. 1992, Judge et al. 1992). Our work attempts to precisely describe the effects of food depletion on shellfish growth and condition index using natural phytoplankton at densities and current speeds that were directly comparable to field conditions. These results were then compared with growth and condition index measurements from animals deployed in the field that were ex- posed to a wide range of resource competition. By comparing the two groups, we were able to make inferences about the food avail- ability conditions experienced by the shellfish in commercial cul- ture conditions over a range of initial stocking densities. These types of measurements are nearly impossible to make directly on a field population because of the challenge of making long-term, continuous measurements of food concentration and current speed. MATERIALS AND METHODS All experiments were conducted in Point Judith Pond in Nar- ragansett. RI (41°24'N; 71°3rW), an 8-km estuary with a 30% semidiurnal tidal exchanae with Block Island Sound (Licata 1981) 4I°25 71° 32 Figure 1. Location map of Point Judith Pond. Narragansett, RL sliowing I A) the location of Camp Fuller where the flume was con- structed, and the locations of the bottom cages (B and C), 300 and 500 m to the east. (Fig. 1 ). The flume study was conducted m a temporary structure built at Camp Fuller, located on Turner Cove, approximately two- thirds of the way up the estuary on the west shore, adjacent to the main dredged navigational channel into the pond. The field study was conducted concurrently, with experimental cages placed 300 and 500 m to the east. The experiment lasted for 6 wk, from August 24 to October 4, 1992. Water for the flume study was pumped continuously to a head tank with a small submersible pump placed on the bottom (with the intake 15 cm above the bottom) at a depth of 1.3 m mean low water (MLW) under a floating dock. The pump intake was pro- tected with a 3-mm mesh, and coarse particles and zooplankton were removed with a 200-jjim mesh bag filter. The chlorophyll concentration of the natural seston was monitored continuously over the 6-wk trial with a Turner Designs (Sunnyvale, CA) Model lOAU flow-through fluorometer connected to a peristaltic pump sampling the head tank. Discrete subsamples were filtered and analyzed for seston concentration by total dry weight and organic content by combustion (6 h at 450°C). Water temperature was monitored continuously with a Temp-Mentor recorder (Ryan In- struments, Kirkland, WA) in the head tank. Juvenile bay scallops, A. irradians, and eastern oysters, C. virginica. were obtained from Moonstone Oysters, Narragansett, Rl, a commercial shellfish aquaculture firm. Hatchery-reared seedstock had been in culture throughout the summer at a nearby location in Point Judith Pond. The starting height (longest axis) of the oysters averaged 45 ± 5.7 mm (SD), whereas scallops aver- aged 43 ± 3.2 mm (SD). One week before the start of the exper- iment, individual oysters and scallops were marked with numbered plastic tags attached with quick-setting manne epoxy. All exper- imental animals were scrubbed to remove epiphytes, toweled dry, weighed to the nearest 0.01 g, and measured (longest axis) to the nearest 0. 1 mm with calipers at the start and end of the 6-wk trial. Flume animals were measured again 2 and 4 wk into the experi- ment. A random sample of 25 animals of each species was sacri- ficed at the beginning of the experiment to determine tissue dry weight, shell weight, and starting condition index. Scallop tissues were further dissected to separate gonad, adductor muscle, and viscera, as well as visually examined for gonad color, a subjective measure of sexual competence. At the conclusion of the 6-wk trial, all experimental animals were similarly sacrificed and measured. Flume Experiments Three experimental plexiglass flumes were constructed with seven experimental compartments each, separated by I3-mm- mesh plastic screens to keep the animals in place. Each flume compartment measured 12x13 cm, and the depth was maintained with a standpipe at 6-7 cm. Each compartment was initially stocked with six oysters (mean length, 42.9 mm; mean wet weight, 7.0 g) or three scallops (mean length, 44.7 mm; wet weight, 23.6 g). Daily flume maintenance consisted of quickly draining the flume so that silt and fecal material could be rinsed out. The seawater system was designed to deliver a constant and identical flow of water to each flume. Current speed was measured midstream in the flume with a thermistor probe flowmeter (La- Barbera and Vogel 1976). Every 4—8 d, the readings from the Turner Designs (model 10-AU) flow-through fluorometer were correlated with discrete chlorophyll samples that were filtered, digested in acetone, and read on another Turner Designs (model 10) fluorometer, followed by acidification (Parsons et al. 1984). Changes in the phytoplank- Food-Limited Growth in Juvenile Shellfish 273 ton species composition, as well as fouling on the photocell itselt. can cause gradual shifts in the relationship between the fluores- cence readings from the flow-through fluorometer and chloro- phyll-a concentrations (Chl-a) determined by acetone digestion. The calibration of in vivo fluorescence readings to Chl-a concen- trations was achieved by drawing samples from representative downstream flume compartments to obtain a range of Chl-a con- centrations. These samples were read with the continuous fluo- rometer, and subsamples were filtered for Chl-a analysis. Calibra- tion curves were constructed by correlating the fluorescence read- ings from the Turner 10-AU tluorometer to Chl-a determinations of the discrete filtered samples (>0.35 |jim). Fouling on the flow- through photocell of the fluorometer was cleaned daily by 10 min of acetone immersion followed by a freshwater rinse and weekly by disassembling and scrubbing the photocell. Additional subsam- ples were filtered (Whatman GF/C) for the determination of seston dry weight and percent organic content (by combustion at 450°C for 6 h). Over the course of the 6-wk trial, 17 feeding experiments were conducted to monitor downstream food depletion in each tlume. This was accomplished by setting the flow-through fluorometer to record 20-s average fluorescence data and by placing the peristaltic pump intake into a downstream corner of each flume compartment until the readings stabilized for at least 1 min. The return flow from the peristaltic pump was always directed into the same flume as the intake, but at a different depth. The sequence of sampling was always from the end of the flume to the beginning so that if the sampling disturbed the filtering activity of animals in a compart- ment, the upstream animals would not be affected. Two experimental trials were designed to examine the possi- bility that natural particle sinking and settlement might have an effect on downstream food depletion in the absence of filter feed- ers. One flume was filled with the empty shells of sacrificed ex- perimental oysters, and a second was left empty. No downstream reduction in Chl-a was noted in either flume. Assimilation effi- ciency was calculated by use of the ash ratio method described by Conover ( 1966). Samples of seston and feces were collected on preweighed, precombusted (450°C) Whatman GF/C filters. Filters were rinsed with isotonic ammonmm formate, dried at 60°C for 48 h, then weighed to the nearest 0.0001 g with a Mettler AE 200, combusted at 450°C, and weighed again. Assimilation efficiency (AEVc) and calculated as: AE% = [100 X (F - E)]/[(l - E) X F] where F is the organic content (percent) of the food and E is the organic content (percent) of the feces. Field Grow-Out Experiments Concurrent with the tlume study, field experiments were con- ducted using the facilities of a nearby commercial aquaculture firm. Moonstone Oysters, in Narragansett, RI. Moonstone uses a modified rack-and-bag shellfish culture method fully described in Rheault and Rice (1995). Shellfish were held in 13-mm-mesh plastic bags (0.6 x 0.6 m) on shelves in cages. The cages rest on the sediment at a depth of 2-3 m, holding the bags 10-60 cm above the pond bottom. Two experimental cages were placed 300 and 500 m to the east of the flume study site (see Fig. I). Ten mesh bags were stocked with 20 or more tagged, experi- mental oysters, whereas additional untagged oysters of a similar size were added to make up the desired initial stocking density. Ten bags were stocked with oysters over a range of initial stocking volumes 1-6 1/bag). For comparison, an additional 10 bags of oysters from the 1991 year-class were stocked over a similar range of starting weights; however, these animals were not measured for individual growth rates or condition index. The total weight of each bag was recorded at the start and end of the experiment. The growth of individual tagged oysters in experimental bags was monitored by noting each individual's total wet weight and length at the start and end of the 6-wk study . At the end of the study, all tagged individuals were sacrificed for the determination of tissue dry weight, shell weight, and condition index. Tidal current speed was measured with a thermistor probe (LaBarbera and Vogel 1976) both inside and outside the grow-out bags to determine the effects of the cage on the local small-scale hydro- dynamics. Data Analysis A Turner 10-AU fluorometer was set to collect 20-min average fluorescence readings throughout the 6-wk trial. There were sev- eral short periods where the fluorometer was offline for download- ing data or cleaning the photocell. Missing data were interpolated from corresponding tidal phases of the previous or following day. Tidal current velocity was modeled by fitting a sine function with an amplitude equal to the maximum measured current speed and a period of 12.4 h. Seston flux for the field study was calculated by multiplying the modeled average tidal velocity by the measured chlorophyll concentration for each 20-min period for the 6-wk study. Predicted tidal heights published by NOAA (Rockville, MD) were compared with locally observed times of tidal maxima or minima to determine the lag time for our location in the pond. The seston flux values for all of the 20-min periods each day were averaged to obtain daily seston flux estimates for the field site. The results of the 17 feeding trials were used to calculate the percentage of available food removed by each individual in the flume. Because this percentage was observed to be relatively con- stant over a wide range of concentrations (see Results), it was possible to estimate the ration available to each downstream flume chamber and, conversely, to estimate the ration consumed by each animal in the flume, every 20 min for the 6-wk trial. Individual daily ratios were calculated by multiplying the percent consumed by the average of 72 20-min average chlorophyll concentration measurements made each day. Condition index (CI) was calculated by use of the methods described by Lucas and Beninger ( 1985): Oyster CI = (100 x TDW)/(WW - SW) and. Scallop CI = (100 x TDW)/(SH) when TDW is tissue dry weight in grams, WW is whole wet weight in grams, SW is weight of shell in grams, and SH is shell height in centimeters. Instantaneous growth, expressed as a per- cent increase in length or weight per day, was calculated as: % incr per day = ln((W,/W„)/t] x 100 when W„ is the initial wet weight in grams (or shell length in centimeters), and W, is the weight (or length) at time t in days. Approximately 10% of the oysters were found to have uniden- tified juvenile oyster disease (Davis and Barber 1994) or were severely distressed by infestation of Polydora websteri (mud blis- ter) (Wargo and Ford 1993). These animals would continue to feed 274 Rheault and Rice R', imally but had abnormal shell growth and condition indices that vvere usually statistical outliers. Three oysters in the field trial suffered shell deformations when their shell grew into the mesh of the bag. A similar number of scallops in the flume ceased growing when their hinge ligament became disarticulated. Animals in the flume experiment that were not growing after the first 2 wk were replaced by healthy individuals. Distressed or obviously infected animals were eliminated from the statistical analysis. Shellfish growth responses to downstream food depletion in the flume study were broken down into three 2-wk subsets and after tests for normality of error terms and equality of variance were analyzed with a multiple analysis of variance (SPSS, Chicago, IL) for the effect of flume position and date. The effect of flume position on condition index (determined only at the end of the experiment) was analyzed with procNPARlWAY (SAS; SAS In- stitute, Gary, NO, and differences in growth and condition index between pairs of ration levels were examined with the Bonferroni (Dunn) T-test (SAS). In the field study, individual responses in growth or condition index were compared over the range of initial stocking densities by linear regression analysis (Statgraphics; Manugistics Inc., Rock- ville, MD). The effect of initial stocking density on the growth rates of entire bags (as percent increase in weight per day) were also examined with linear regression. RESULTS There was a strong effect of the tide on both chlorophyll con- centration and temperature (Fig. 2). As the flood tide brought colder, more oligotrophic, ocean waters into the estuary, temper- ature would decline by 2-6°C and chlorophyll concentrations would drop by as much as 80%. Although the temperature varied greatly over a tidal cycle, there was only a 0.6°C difference be- tween the average temperatures over the first two 2-wk sample periods (8/24-9/8 mean, 20.7°C, 9/8-9/21 mean, 20.1°C). The Chlorophyll, Temperature, Tidal height Figure 2. A representative sample of continuous data collected during the 6-wk flume experiment. Chl-a (jig I') and temperature (°C) are plotted with tidal height (m). third 2-wk sample period, 9/21-10/2, showed a three-degree de- crease in the average temperature to 16.7°C. Chl-a concentrations were similarly variable with the tides (Fig. 2; Appendix III in Rheault 1995), but daily averages of chlorophyll concentrations were relatively stable, between 4 and 7 jjLg 1 ~ ' for most of the experiment. Two large rain storms on 9/2-3 and 9/27 caused Chl-a concentrations to decrease sharply (Fig. 3b). Chl-a concentrations averaged 5.7 |jig/l for the first 2 wk, 4.6 (jLg 1 ' for the second 2 wk, and 4.5 (xg • ' for the last 2 wk. Size fractionation of the seston revealed that the majority of the seston was smaller than 20 (i-m: 78 ± 11% of the total seston (SD; n = 17), 70 ± 15% of the organic component of the seston (SD; n = 17) and 78 ± 8% of the chlorophyll (SD; n = 8). Seston concen- trations were highest and had a greater percentage of organic mat- ter in the first 2 wk of the experiment (Fig. 3a). Flow rates to each flume were checked daily for the first week and twice a week thereafter. For the first 4 d, the flow rates were 1.7 to 1.9 I min '. To increase the degree of downstream food depletion, we decided to slow flow rates to I.O-I.l 1 min"' for the next 4 wk. In the last week of the experiment, ambient Chl-a concentrations were depressed after a 7-cm rainstorm, so tlow rates were increased to 1 .5 1 min" '. Seston flux to each flume was 2" rain - TSES • %OM 60% 50% ; 40% 30% [ 20% o 'c nj c 0) o 0) Q. 3' rain average chlorophyll (ug/liter) flume flow rate (liter/m) avg=20 I '"WWAV*^\JSAV^ 'n avg=16 7 \ry Figure 3. The graphs describe food, flow, and temperature conditions in the flumes over the course of the 6-vvk experiment (8/24/92 to 10/ 2/92). (a) Discrete samples of total seston concentration are indicated by triangles, and percent organic matter ( % OM) in those seston sam- ples is indicated by circles, lb) Daily average Chl-a concentration cal- culated from a continuous record of 20-min means. Arrows indicate two rainstorms that noticeably diluted the Chl-a concentration. Flow rate of water in the flumes is given, (c) The temperature record (°C) is 20-min averaged data recorded continuously in the head tank feeding (he flumes. Average temperature values for each 2-wk period are indicated. Food-Limited Growth in Juvenile Shellfish 275 calculated as the product of the average Chl-a concentration and the flow rate to each tlume (Fig. 3b). Current speed at the center of the flume was measured at 0.38 cm/s at a flow rate of l.I I min ^ ' . Downstream food depletion was calculated 17 times over the 6-wk trial by measuring the decline in Chl-a concentration in each compartment and was expressed as a percentage of the incoming concentration. Food depletion at each stage of the flume remained stable despite fivefold differences in incoming Chl-a concentra- tion, although higher flow rates at the beginning and end of the experiment did reduce depletion percentages. Clearance rates were depressed in the last week of the experiment when temperatures dropped below I6°C (Fig. 3c). The feeding rate, or ration con- sumed (as percentage of incoming concentration), for each tlume compartment was calculated from the difference in Chl-a concen- trations between compartments. The downstream Chl-a depletion percentages from 17 feeding experiments were averaged for each flume compartment, and an exponential depletion function was fit to the data (Fig. 4). The best-fit downstream depletion curves for the oysters show that each tlume compartment removed 27'7f of the food flowing in from the chamber above whereas scallops removed 35%. The daily ration (Fig. 5) was calculated by multiplying the daily average chlorophyll concentration (Fig. 3b) times the tlow rate (Fig. 3b) times the percentage consumed by each compart- ment (Fig. 4). The feeding studies indicate that oysters in the first flume compartment consumed an average of 0,38 mg of Chi d^ ' per oyster, and scallops in the first compartment consumed an average of 0.99 mg of Chi d" ' per scallop. Downstream animals consumed proportionally smaller amounts. For comparison pur- poses, it is convenient to express these rations on a per gram of tissue dry weight basis. Using average starting tissue dry weights of 0.37 g per oyster and 1.33 g per scallop, an oyster in the first flume compartment consumed a ration of 1 .03 mg of Chi d ~ ' per g dry wt, and a scallop consumed 0.74 mg of Chi d" ' per g dry wt. There was no consistent trend in the percent organic matter in either the downstream seston or feces samples for either scallops or oysters (Table 1). Scallops had a consistently higher assimila- tion efficiency (AE%) than oysters, whereas AE% declined for both species as the organic fraction of the seston declined later in the experiment. Flume Growth Both oysters and scallops responded to decreasing downstream food availability with similar declines in incremental growth (per- cent per day increase in both length and weight) (Figs. 6 and 7) as well as condition index (Fig. 8). Incremental growth rates also declined with time over the course of the 6-wk experiment. Growth rate (both length and weight increases over the entire 6-wk trial) and condition index (determined at the end of the experiment) were significantly correlated with the calculated av- erage ration for each flume compartment (Tables 2 and 3). Pair- wise analysis of the condition indices of oysters from each com- partment suggests that oysters can respond measurably to reduc- tions in ambient food concentration as low as 27% of the initial concentration. The condition index was the only measure that detected significant differences between the first two compart- ments (p = 0.0025). Percent increases in length or weight were more variable and did not show significant differences between Downstream Chlorophyll Depletion best fit - oysters Downstream Chlorophyll Depletion best fit - scallops 100%© percent remaining after each compartment 2 3 4 5 6 7 flume compartment Figure 4. These plots show dounstreani chlorophyll depletion in the flume as the percentage of the incoming concentration remaining at each flume compartment for oysters (a) and scallops (b). Abscissa numbers refer to flume compartment numbers. Each point is the av- erage of 17 feeding experiments. Semilogarithmic transformation of these Chl-a depletion data showed that scallops removed a greater percentage of available food (35% at each compartment) than did oysters (27% at each compartment). The ration consumed is the dif- ference in concentration between compartments. adjacent chambers, even though the correlations with ration were highly significant (Tables 2 and 3). Field Grow-out Study The field grow-out study showed the growth responses of oys- ters to variations in initial stocking density measured three differ- ent ways: change in weight of the entire bag. individual growth in total wet weight, and individual condition index. Bags were ini- tially stocked at 1-5 kg/bag (equivalent to 2.7-13.5 kg/m~-). The low-density bags more than doubled in weight in both the 1991 and 1992 year-classes, but smaller 1992 year-class oysters grew faster than the 2-yr-old 1991 year-class (Fig. 9). Individual whole wet weight increase (percent per day) was negatively correlated with starting bag weight (/•- = 0. 128. p < 0.0001 ). Oysters in the lightly stocked bags increased weight at an average 2.4% per day. compared with only 1 .9% per day in the heavier bags (Fig. 10a). The condition index for tagged individuals was also negatively 276 Rheault and Rice TABLE 1. Percent organic matter in seston and feces and assimilation efficiency. (AE%) Date Flume Chamber Number Seston, % Organic Feces, % Organic Oysters Scallops Sept. 1 Sept. 3 Sept. 5 Sept. 18 35.0 37.4 32.6 15.6 ND* 19.4 18.5 12.7 15.4 Mean AE% 1 4 7 35.0 49.9 48.5 52.6 17.5 60.7 25.5 24.0 28.2 15.3 66.5 21.7 ND 23.1 Mean AE% 1 4 7 50.0 35.7 32.1 38.6 25.8 65.3 19.3 30.7 23.8 22.4 71.2 23.8 22.5 25.0 Mean AE% I 4 7 36.2 25.5 28.0 21.4 24.6 42.4 ND ND 18.3 23.2 46.8 ND ND 18.9 Mean AE% 25.0 18.3 32.7 18.9 30.2 * ND, not done. correlated with starting bag weight (R' = 0.321. p < 0.0001) (Fig. 10b). The average condition index varied from a high of 7.6 in the lightly stocked bags to 5.4 in the heavily stocked bags. Maximum tidal velocities (measured at peak flood tide) were 5-8 cm/s, and the calculated average current speed was 4. 1 cm/s. Current speeds measured inside the mesh bags averaged 10% of the bulk flow measured near the experimental cages. The horizon- tal chlorophyll flux in the field had greater short-term variability than did the flux to the flume because the field flux had the tidal velocity variations superimposed on the variation in concentration, whereas the flume experienced a constant flow rate. However, daily averages of the 20-min flux estimates were nearly identical to the flux estimates in the flume (Fig. 11). DISCUSSION Several workers have studied feeding in scallops using race- ways and flume. Kirby-Smith (1972) recommended that to main- tain maximal growth of bay scallops in a raceway system, the effluent water should contain at least 60% of the incoming phy- toplankton concentration. Rhodes et al. (1981) calculated that chlorophyll concentrations above 1.0 (jig 1 ' are enough to main- tain maximal growth in bay scallops. The data presented in our study suggest that growth rate and condition index will decline after even smaller decreases in food concentrations. Significant reductions in oyster condition index were detected in the second fliiiiie chamber where chlorophyll concentration was reduced by only 27%. Cahalan et al. (1989) used flume experiments to try to separate the effects of current speed and food concentration on growth in bay scallops, but their study used cultured algae and low popula- tion densities and did not result in food depletion. Similar studies with scallops feeding on natural particulate by Kirby-Smith ( 1972) and Kirby-Smith and Barber (1974) showed food-limited growth at slow current speeds but failed to describe the relationship be- tween available ration and growth. Rhodes et al. (1981) studied the carrying capacity for bay scallop seed in raceways and pro- jected a minimum ration for maximal growth of 74 nig of Chi d ~ ' per liter of biomass. Scallops of the size used in our study would pack 41 to the liter and collectively consume an average of 40 mg of Chi per day. Rhodes' scallop seed were considerably smaller (<10 mm) than those used in this study (mean, 44 mm). Smaller scallops will pack tighter (more dry weight per unit volume), and smaller shellfish in general will consume more on a per gram dry weight basis (Dame 1972). Seston Flux Weekly phytoplankton counts over the summer and fall of 1990 in Point Judith Pond showed that Skelelonema cosiarum was the numerically dominant species and that chlorophyll concentrations were consistently 2-10 times greater at the head of the pond than at the mouth (Rheault 1995). suggesting that the pond is a net source, rather than a net sink, of primary productivity. In this study, the tidal exchange resulted in two- to fivefold variations in chlorophyll concentrations, peaking at low tide and declining until high tide (Fig. 2). This observation has important implications for studies that rely on single daily or weekly seston sampling to estimate food availability. One can minimize the sampling error by consistently sampling at the same phase of the tide, but errors of a few hours either way could have drastic effects on estimates of average food concentrations. This phenomenon will be most pronounced in tidal estuaries (such as Point Judith Pond) where rich eutrophic waters are mixed with a salt wedge of oligotrophic ocean waters. Like- wise, simply sampling surface waters may grossly overestimate Daily Ration Consumed avg. mg Chl/day per animal 1.4 .1.2 Q (0 . 1.0 + to.8 o0.6 O) ^0.4 d) > "0.2 0.0 6 12 3 4 5 flume compartment Figure 5. The average daily ration consumed is the product of Chl-a concentration (p.g I '; Fig. 3b), the flow rate (1 min '; Fig. 3b), and the percentage of food removed by each compartment (Fig. 4), divided by the number of animals in each compartment (six oysters or three scallops). Daily chlorophyll ration is expressed in units of mg d ' per scallop (a) or per oyster (b). Data are presented as the mean value plus or minus 1 SD for the 6-wk flume trial. Food-Limited Growth in Juvenile Shellfish 277 z < 1 6 z t— I O z < Q-04 Oysters 0.0 --+--^1 __+"i f \ \ .----"'~~-^ \l 1 _L_ J_ _L J 3 15 - 27 6 39 1 2 DATE - Figure 6. Instantaneous oyster growth is presented as percent in- crease per day in weight (a) or length (b) for each of the flume com- partments. The mean growth rates (n = 10 or 12) for each 2-wk period are plotted separately, with standard error (S.E.) bars. The top curve represents growth in the first 2 wk, the middle curve is growth in the second 2 wk, and the bottom curve represents growth during the last 2 wk. and calculated an average current speed of 4.1 cm s~ '. Current speeds inside the mesh bags were measured directly at 10% of the bulk now, or 0.41 cm s" '. For comparison, the current speed in the center of the flume measured 0.38 cm s" ' when the flow rate was 1.11 min' ', suggesting that (on average) the flume animals were experiencing conditions very similar to those of the field animals. However, the flume animals were exposed to a constant current speed, whereas the field animals experienced an oscillating current regime. The variation in current speed caused increased short-term variability to the flux estimates; however, daily aver- ages of the flux to the flume and the flux to the field were not significantly different. Feeding Loosanoff (1958) found no effect of temperature on pumping rate in adult oysters taken from Long Island Sound at temperatures between 16.0 and 28.0°C; however, below 16.0°C, he noted a 50% decrease in pumping rate. Other researchers have subse- quently modeled bivalve clearance rate response to changes in temperature (Doering and Oviatt 1986) and found a nearly linear response. Bayne et al. ( 1977) also reported a direct effect of tem- u z h- 2 I O < z LU Scallops food conditions experienced by bottom-dwelling suspension feed- ers. When the water column is stratified the benthos will be bathed in the cool, saline, dense, oligotrophic ocean waters, whereas the surface waters are relatively warm and food rich. Furthermore, these shifts in food concentration and water temperature are at odds with the description of Point Judith Pond as a typical •well- mixed" estuary (Licata 1981). The semidiurnal fluctuations in both tidal currents and chloro- phyll concentrations make the estimation of field seston flux very difficult. Maximum and minimum chlorophyll concentrations were observed at low and high tide, respectively, periods when the tidal current speed drops to zero. An additional confounding factor was the fact that current speeds in the shallow ( <2-m) pond can be greatly influenced by wind speed and direction. We have observed a number of occasions when moderate breezes would cause local water currents to move in the opposite direction of the tidal flow. The only way to get an accurate estimate of seston flux in these field conditions would be to place a continuously recording current meter on site with the continuously recording fluorometer. Lacking these data, we estimated field fluxes from a rudimen- tary sine wave model of the cuirent speed (Rheault 1995, Appen- dix 111). We measured maximum current speeds of 5-8 cm s" ' 08 r z X 0.6 O z LU ^0.4 >■ < d ^02 00 :^-^ j_ 15 28 6 •40 DATE — Figure 7. Instantaneous scallop growth is presented as percent in- crease per day in weight (a) or length (b) for each of the flume com- partments. The mean growth rates (n = i) for each 2-wk period are plotted separately, with standard error (S.E.) bars. The top curve represents growth in the first 2 wk, the middle curve is growth in the second 2 wk, and the bottom curve represents growth during the last 2 wk. 278 Rheault and Rice CO 0) -n c c O c o u 0 O) D 4 - - 3 > < _ \ - 1 i OYSTER 1 1 ' ^ 4 1 1 - 4 • SCALLOP 1 1 1 • 1 1 Chamber r I ume Figure 8. Mean oyster (curve A) condition index (± standard error, S.E.) at the end of tlie 6-wk flume experiment is plotted for each flume compartment (n = 10 or 12). Mean scallop condition index (±S.E.) after the trials is plotted in curve B (n = 3). perature on both filtration rate and assimilation efficiency in mus- sels. Temperatures for the first 4 wk of our study remained well above Loosanoff's 16.0°C threshold but averaged only 16.7°C for the last 2-wk period (Fig. 3c). Thus, it is likely that temperature declines in the last 2 wk accounted for much of the observed decline in growth rates for both the scallops and the oysters over this time. We observed that the percentage of food removed at each stage of the flume remained stable despite fivefold differences in incom- ing chlorophyll concentration. At the beginning of the experiment, when flow rates were slightly elevated, the percentage removed at each stage was correspondingly smaller. These two findmgs imply that the shellfish were clearing a fixed volume of water regardless of the food concentration. This conclusion is supported by Haven and Morales-Alamo (1970), who also described a "well defined pattern of particle removal (by oysters] when results are expressed in terms of percent removal." Tenore and Dunstan (1973) also reported that oysters removed a constant percentage of the food from the water when the ambient food concentration was within the range of 0.26-0.76 mg of C per liter, a "food concentration typical of natural environments." Below these levels, they re- ported sharply lower clearance rates. Similarly, Myiilus edulis has been observed to cease feeding if food concentrations drop below a certain concentration, a so-called "threshold feeding response" (Ihompson and Bayne 1972, Wilson and Seed 1974, Bayne and Scuilard 1976, Butman et al. 1994). This concentration may cor- respond 10 the level where the energy required to pump water past the gills exceeds the energy gained by the food captured. During our experiment, the incoming seston concentrations were always in Tenore and Dunstan's (1973) "typical" range; however, downstream concentrations dropped off rapidly. Head tank seston concentration averaged 1.36 mg P' ash free dry weight (AFDW). corresponding to 0.45 mg of C per liter (assum- ing a ratio of 0.33; 1 carbon;AFDW. calculated from Prosser 1973). Seston concentrations in the flumes declined rapidly down- stream and may have regularly dropped below the concentration where animals could reap an energetic benefit by active feeding. This would explain why scallops in the last three compartments consistently removed less chlorophyll than was predicted by the curve fit to the data from the feeding rates of the scallops in the first four compartments (Fig. 4). To our knowledge, this would be the first report of a "threshold feeding response" in A. irnidians irradians. An alternative explanation for this could be that any chlorophyll remaining in the flume after the fourth compartment was composed of particles that were too small to be effectively retained by the scallops but were still detected by the fluorometer and the individually filtered samples. Palmer (1980) reported that bay scallops adjust their clearance rate according to ambient food concentrations to ingest algae at a near-constant rate. This conflicts with our observations that scal- lops cleared a near-constant percentage of the available food over a fourfold range of Chl-a concentrations. In our flume experi- ments, downstream scallops filtered a slightly smaller percentage of the available ration in contrast to Palmer's predicted response of increased clearance rate at lower food concentrations. There was no consistent trend in the percent organic matter in either the downstream seston or the feces samples for either scal- lops or oysters (Table 1). Any decline in the percentage of organic matter in the remaining downstream seston would be an indication that the shellfish were either selectively filtering organic particles and leaving inorganic silt in suspension or perhaps feeding on resuspended fecal material (Loosanoff 1949. Newell and Jordan 1983). Effect of stocking density on oyster growth (whole bag weight) 2.8% - 2.4% I- *92 year class Y = - 0.0013.\ + 0.0264 R squared = 0.772 '91 year class Y=-0.0019.\ + 0.0222 R squared -0.814 10 20 30 40 initial stocking weight (kg/bag) 50 Figure 9. The effect of varying initial stocking density (kg per bag) on oyster growth in the field study. Weight increases of entire bags of oysters are presented as percent increase per day. Six-month-old oys- ters from the 1992 year-class (filled squares) grew faster than larger I8-month-old oysters from the 1991 year-class (hatched squares). Cal- culated linear regressions are plotted and formulae are given for each year-class. Both regressions were significant (p < 0.0005), and the slopes were significantly different from zero (p < 0.0001). Food-Limited Growth in Juvenile Shellfish 279 TABLE 2. Statistical analyses of growth and condition index response. Vs. Condition Index Vs. % Incr/d Weight Vs. % Incr/d Length Oysters Linear regressions of ration (mg of Chl-a/d/oyster) y-intercept Standard error (SE) of y estimate No. of observations Degrees of freedom (n - 2) X coefficient SE of coefficient Probability > F Scallops Linear regressions of ration (mg of Clil-a/d/scallop) y-intercept SE of y estimate No. of observations Degrees of freedom ( n X coefficient SE of coefficient Probability > F 2) 3.29 0.754 0.839 75 15.98 0.8201 0.0001 1.57 0.353 0.889 18 16 3.26 0.2878 0.0001 0.93% 0.32% 0.31% 0.14% 0.607 0.361 75 75 73 73 3.59% 1.01% 0.0034% 0.0016% 0.004 0.0001 1.43% 5.29% 6.06% 3.04% 0.942 0.817 18 18 16 16 79.50% 20.93% 0.0494% 0.0248% 0.0001 0.0001 Clearance Rate The flume study data can also be used to estimate individual clearance rates if one assumes KlOVr retention efficiency on the gill and no refiltering within each compartment. Six oysters cleared 21% of the food available, with average flow rates of 1 . 1 1 m~'. This is equivalent to a 50 ml/min clearance rate for a 45-mm oyster with an average tissue dry weight of 0.40 g (or 124 ml min" ' per g dry wt). This agrees well with Jorgensen's ( 1966) estimate of 129-208 ml min" ' per g dry wt but is four times the rate reported by Palmer (1980) for much larger (60- to 100-mm) oysters. Dame et al. (1984) reported ingestion rates of oysters feeding on natural particulate of 0.39-2.02 mg of Chi d" ' per g dry weight, which agrees well with the average value of 0.74 mg of Chi d ~ ' per g dry weight reported here. Three scallops cleared an average 35% or 128 ml min" ' (equivalent to 97 ml min~ ' per g dry weight for an average 43-mm scallop with a tissue dry weight of 1.3 g). This agrees well with Palmer's (1980) average clearance rate of 95 ml min " ' per g dry weight reported for TABLE 3. For oysters, Bonferroni (Dunn) pairwi.se T'-tests for significant difference between mean growth and condition index from each flume compartment. Different letters denote significant differences between means (a = 0.05). ' wet algal weight of cultured Condition Weight Length Flume Position Index % incr/d % incr/d 1 a a a 2 b a a 3 c ab ab 4 cd b b 5 de b b 6 de c c 7 de c c 40-mm scallops fed ffl 3.1% 1.1% T3 C 1 ' ■ ' ■ 1 ■ I ' ' -. — . — 1 — . — . — . — .— ] ** - ~ .. * 1 • * :^^'i^ t -* ^^^■^" : - ~ _ _ * 1 ■Y=-0.12%X + Z5% ~ ~ ~ * . p<0.0005 - ~ -^ ~ • " , ..... 1 . 1 1 1 0 12 3 4 5 Initial bag weight (kg) Individual CI vs. Bag Weight 3.6 P Y=-0.60X+8.21 p<0.0001 0 12 3 4 5 Initial Bag Weight Figure 10. Individually tagged oysters from bags (Fig. 9) were mea- sured after the 6-wk field study. Growth rates (percent increase in weight per day [a]) and condition index (b) are plotted against the initial stocking density of the bag. Linear regression formulae are shown with each line plotted with 95% confidence limits (inner pair of dashed lines) and 95% prediction limits (outer dashed lines). last 2-wk period. Unfortunately, because both temperature and chlorophyll concentrations were declining over this period, our experimental design does not allow us to separate out the effects of these two variables. Field Study Extrapolations The calculated average current speed in the field grow-out bags (0.41 cm s ') was nearly identical to that measured in the flume (0.38 cm s '); however, the field animals experienced an oscil- lating tidal current, whereas the flow to the flumes was constant. Thrse oscillations in flow resulted in large fluctuations in flux, such that four times a day, flux decreased to zero and maximum flux rates would be 1.7 times that experienced in the flume. It is uncleai- what the effect of this variability in food availability and flow has on the organism's ability to feed and assimilate food; however, there is some evidence that an intermittent feeding re- gime may enhance growth (Langton and McKay 1976). In spite of the cyclic nature of the current speed, the daily average of all the 20-min flux estimates yielded a daily flux rate similar to that calculated for the flumes (Fig. 1 1 ) (see also Rheault 1995, Appendix Illl. Therefore, by comparing growth rates and condition indexes between the two studies, it is reasonable to infer the degree of seston depletion at various stocking densities in the field. Although there may be differences in the individual animal's perception of local food-depletion effects (depending on whether that individual is located in the middle of the grow-out bag or near the edge), the average growth response and condition index data from individuals in the grow-out bags should be comparable to those from the flume study. Weight increases by individual oysters in the high-density bags ranged from 1.7 to 2.2% d"' (Fig. lOa) and are comparable to those recorded in the second and third flume compartments (Fig. 6a). Oysters in low-density bags (1-2 kg per bag) appeared to grow slightly faster, at 2.4% d " ' , than did those in the first flume compartment, at 2.2% d '. Oyster condition index in the field study ranged from 5 to 8 (Fig. 10b), corresponding to condition indexes in the fifth and second flume compartments, respectively (Fig. 8). Extrapolating the growth response of the field animals to the flume study, we project that the low-density bags were expe- riencing a 27% local seston depletion, whereas the high-density bags were experiencing a 72% reduction in the ambient seston concentration. Seven bags containing tagged scallops were stocked at 1.8 kg per bag. The average condition index was 4.2 ± 0.57 (SD; n = 68), similar to the condition index recorded in the first or second flume compartments. The average length increase was 0.5 ± 0.06% per d"' (SD), which was similar to the second flume compartment, and the average total wet weight increase was 0.75 ± 0.23% d"' (SD). comparable with the third or fourth flume compartments. Overall, scallops stocked at 1.8 kg per bag had a growth response similar to those in the second flume compartment; thus, we can infer that on average they were experiencing local seston concentrations approximately 42% of the ambient seston concentration outside of the bags. Food Quality Several researchers have remarked that chlorophyll concentra- tion is a poor estimate of food availability or food quality for Dally Average Chlorophyll Cone, with Calculated Field Flux 10 c 6 O O 1 4 □. o o avg. cone. • avg. flux. «. -.8 . *i ■ a. 40 ^ o 0) 35 u CM 30 1 o 25 o o 20 1 X 15 3 u. 10 ■D 5 ^ . O 08/24 08/31 09/07 09/14 09/21 09/28 10/05 Figure 11. Daily average Chl-a concentration in flume (squares) is plotted with calculated daily Chl-a flux for the field site (circles). Food-Limited Growth in Juvenile Shellfish 281 suspension-feeding bivalves (Soniat et al. 1984, Wikfors personal communication). The in vivo method of measuring chlorophyll has its own limitations because changes in species composition, light adaptation, or nutrition can influence the phytoplankton ex- citation-emission response (Loftus and Seliger 1975). Nonethe- less, earlier work in Point Judith Pond (Rheault and Rice, ac- cepted) examining the food-limited growth of three species of juvenile shellfish demonstrates that the fluxes of both chlorophyll and particulate organic matter (POM) were significantly correlated with shellfish growth. Stepwise multiple regressions of chloro- phyll or POM nux and temperature accounted for over S59c of the variance in growth. The continuous recording tluorometer permitted us to accu- rately chronicle the temporal variability in food availability char- acteristic of estuarine environments. It is questionable whether any discrete sampling regime could adequately characterize the food concentration under these conditions. Whereas the measurement of carbon, nitrogen, lipid, protein, carbohydrate, or caloric con- tent might have provided more physiologically meaningful esti- mates of food, these measurements can only be made on discrete samples and do not readily give one a picture of the true variability of these features in a dynamic estuarine environment. Indices of Growth This study used several approaches to monitoring the growth and condition index responses to varying conditions of food hm- itation. For oysters, the condition index was determined using the ratio of dry tissue weight to the estimated shell cavity volume (total wet weight to shell weight) (Lawrence and Scott 1982). This approach avoids some of the problems associated with varied shell thickness and morphology in oysters. We also monitored growth in shell height (longest axis) and total wet weight. Shell height showed the most variability and had the poorest correlation with ration, flume position, or bag stocking density. Condition index proved to be the most sensitive of the indices to changes in ration downstream and is the preferred method of assessing the health of a population (Lucas and Beninger 1985. Rainer and Mann 1992). For scallops, we used a simpler condition index, based on the ratio of tissue dry weight to shell height (longest axis) because scallop shell morphology is more regular than that of oysters. The variability in shell growth rate and condition index was much less than that seen in wet weight growth, possibly because a small percentage of the scallops were visibly gravid. It was not antici- pated that 6-month-old animals should be capable of developing ripe gonads (Barber and Blake 1981). For oysters, condition index was the most consistent static index of physiological state (Lucas and Beninger 1985), showing the lowest coefficient of variation and the best correlations with ration or bag-stocking density. Condition index provides a reliable measurement for comparing similar-sized animals; however, con- dition index varies with size, season, and reproductive output, as well as with physiological state. For condition index to become useful as a spot check index of physiological state would require data on well-fed and starved animals over a range of sizes and seasons. With these data, one would have a reference point with which to instantly gauge the health of the sample in question (Lawrence and Scott 1982). CONCLUSIONS In conclusion, the downstream depletion of natural seston in a flow-through flume resulted in marked reductions in growth and condition index. Chlorophyll concentration was monitored contin- uously in the incoming flow, and downstream food depletion was characterized several times over the 6-wk trial. The bivalves in each stage of the flume removed a constant percentage of the available food over a wide range of concentrations, resulting in smaller rations downstream. Scallop and oyster clearance rates (milliliters per minute) were constant over a wide range of chlo- rophyll concentrations, suggesting that these species will filter natural seston at a near-constant rate, despite fourfold tidal vari- ations in food concentrations. Scallop clearance rates were re- duced when chlorophyll concentrations were depleted to below 12*^ of the natural levels, suggesting a threshold feeding response similar to that reported for mussels (Thompson and Bayne 1972. Wilson and Seed 1974). Concurrent field grow-out studies con- ducted nearby compared the effect of varying initial stocking den- sities in a commercial shellfish aquaculture operation. Comparable changes in growth and condition index were observed in the field trials, allowing estimates of in situ horizontal seston flux. Al- though physiological responses to food limitation will necessarily be site specific to varying combinations of temperature, current speed, and food concentration or quality (Newell and Shumway 1993). this work provides an opportunity to compare the growth response of oysters and scallops under a wide range of food avail- ability in both laboratory and commercial aquaculture settings. In these experiments, doubling the stocking density from 2.5 to 5.0 kg of oysters per bag resulted in a 209c decrease in both the condition index and the growth rate (percent increase in weight). These observations may provide valuable insights to the commer- cial grower by assisting in decisions pertaining to optimal stocking density for aquaculture grow-out systems. ACKNOWLEDGMENTS This work would not have been possible without the coopera- tion and generosity of many people. We are especially grateful to the YMCA and Gary Richardson, the Director of Camp Fuller, for permitting us to construct a temporary shelter on their beach to conduct the flume experiments. We are deeply indebted to Dr. Charles Roman of the National Park Service for the loan of the Turner 10-AU continuous recording fluorometer and to John Karlsson. Rl DEM, for the use of the Temp-mentor. Special thanks are due to Christian Vye for his assistance in the statistical analyses. We also thank Bob Bergen and the employees of Moon- stone Oysters for their assistance and patience during these exper- iments. Thanks also to Ann Rheault, Dr. John McN. Siebruth, Dr. Candace Oviatt. and Dr. David Bengtson for their assistance in editing the manuscript. This is publication number 3077 of the Rhode Island Agricultural Experiment Station. LITERATURE CITED Asmus. R. M. & H. Asmus. 199L Mussel beds: limiting or promoting phytoplankton? y. Exp. Mar. Biol. Ecol. 148:215-232. Barber. B. J. & N. J. Blake. 1981. Energy storage and utilization in relation to gametogenesis in Argopecten irradians conceniricus (Say). J. Exp. Mar. Biol. Ecol. 52:121-134 Bayne. B. L. & C. S. Scullard. 1976. An apparent specific dynamic ac- tion in Mytilus edulis L. J. Mar. Biol. Assoc. U.K. il:l>l\-il^. Bayne. B. L.. J. Widdows & C. W. Worrall. 1977. Some temperature 282 Rheault and Rice relationships in the physiology of two ecologically distinct bivalve populations. In: F. J- Vernberg. A. Calabrese. F. P. Thurberg and W. Vemberg (Eds.). Physiological Responses of Marine Biota to Pollut- ants. Academic Press, New York. pp. 379-400. Butman. C. A.. M. Frechette. W. R. Geyer & V. R. Starczak. 1994. Flume experiments on food supply to the blue mussel Mytiliis edi 0) if 40 (0 ■o re c o O c re 0) 30 20 10 r ■ C. virginica ■ C. gigas 1 ^11 T .mi J Jll T IVIar Apr IVIay Jun Jul Aug Month -1992 Sep Oct 1400 1200 (0 1000 < u o o c re 0) 800 600 400 200 X ■ C virginica ■ C gigas I ■ J-i^T. rji Jr III diil 1 II Mar Apr May Jun Jul Aug Sep Oct Month -1992 Figure 4. Mean ( -I- 1 SD) oocyte areas of C. virginica and C. gigas from March through October 1992. decrease in mean oocyte area of 700-800 (xm". as the larger, more mature oocytes were released during spawning. There were no female oysters with measurable oocytes in September or October. As indicated by two-way ANOVA. there was a significant difference in mean oocyte area between months (P ^ 0.001) but not between species (P > 0.122). There was also a significant month X species interaction (P =s 0.001). again supporting the existence of different patterns of gametogenesis between oyster species, as determined by oocyte area. Even though there was not an overall difference in mean oocyte area between species, the mean oocyte area of mature (Stage 3) females was significantly greater (P ss 0.001 ) for C. gigas. averaging 1 ,236 |xm". compared to 820 |jLm" for C. virginica. There was a considerable difference in the prevalence of par- asites infecting the two oyster species. Parasites were only found in C. virginica (Table 1 ). The combined prevalence of infection by H. nelsoni and P. marinus ranged from 0% in March through May to 86.7% in August; prevalence decreased to 26.7% in October. The intensity of infection by both parasites, as indicated by the number of systemic infections, increased from June through Sep- tember before declining in October (Table 1). Individuals infected by both H. nelsoni and P. marinus were seen in July (n = 1), TABLE 1. Number of oysters, C. virginica, infected by each of the parasites H. nelsoni and P. marinus and overall prevalence ( '?c of oysters infected) on each sampling date. H. nelsoni P. marinus Prevalence Date Oysters U-E-S U-E-S (%) March 17 15 15-0-0 15-0-0 0/15 = 0 Apnl 17 15 15-0-0 15-0-0 0/15 = 0 May 16 15 15-0-0 15-0-0 0/15 = 0 June 17 13 10-2-1 9-4-0 7/13 = 53.8 July 16 15 9-1-5 10-5-0 10/15 = 66.7* Aucust 14 15 6-3-6 5-5-5 13/15 = 86.7* September 18 15 10-0-5 7-3-5 12/15 = 80.0* October 20 15 13-1-1 13-1-1 4/15 = 26.7 Figure 3. Mean (-1-1 SD) GAIs of C. virginica and C. gigas from March through October 1992. Infection intensity is categorized as; U = uninfected; E = epithelial; S systemic. * Includes oysters infected by more than one parasite species. 288 Barber August (n = 6), and September (n = 2). In addition, two indi- viduals (n = I in both June and September) infected witii P. marinus also contained sporocysts of the trematode Bucephalus cuculus. The weekly mean water temperature and salinity in the York River. VA, in 1992 are shown in Figure 5. Temperature generally increased from a low of 3.8°C in February to 27.5°C in August and then steadily decreased from September through December. Sa- linity was generally above 20 ppt, fluctuating between a low of 17.6 ppt and a high of 23.4 ppt. Salinity was below 20 ppt from April to July and again in September. DISCUSSION There were clear differences between the two oyster species examined in this study in both qualitative and quantitative aspects of gametogenesis, even though both experienced identical envi- ronmental conditions. The primary qualitative difference in game- togenesis was the lack of a clearly defined, synchronous, and complete cycle of gamete development and spawning in C. vi- rginica. Although sexual differentiation occurred as early as March and gamete development continued until June, only two individuals having mature gametes were seen and at no time was evidence of spawning observed. Instead, beginning in July, a large proportion of oysters began to resorb gametes, as follicles began to shrink and were infiltrated by hemocytes. In addition, inactive oysters were present in all months. The lack of postspawn indi- viduals and the large number of inactive and resorbing individuals suggest that spawning never occurred in C. virginica. In contrast, C. gigas exhibited a complete, well-defined, and synchronous gametogenic cycle in which a single spawning occurred between 17 June and 16 July. Like C. virginica. sexual differentiation began as early as March and continued through April. The greater number of inactive individuals seen in May compared with April could be related to the relatively low salinity (<20 ppt) occurring at that time; the optimal salinity for the development of C. gigas larvae at 25°C is between 20 and 26 ppt (Amemiya 1928). Gamete development was rapid after 16 May. however, and most individ- uals were mature in June. In July, most individuals had spawned. This was followed by minor redevelopment in August and rapid resorption in August and September. Unlike C. virginica, no in- active C. gigas were seen from June through August. There were significant differences in both GAI and oocyte area over time for both C. virginica and C. gigas. Overall, the mean GAI was sig- 30 25 20 15 10 5 7 / ^ \ ^xy^^^^^ Temperature (°C) Salinity (ppt) 1 JFMAMJ JASOND Month -1992 Figure 5. Weekly means of water temperature and salinity in the York River, VA, in 1992. nificantly greater for C. gigas than for C. virginica. but the mean oocyte area was similar for both species. The fact that there was a significant mteraction between species and month for both GAI and oocyte area is reflective of the fact that the pattern (or extent) of gametogenesis, as measured by these two parameters, differed for the two oyster species. There were major differences in disease prevalence between C. virginica and C. gigas. In the case of C. virginica. neither H. nelsoni nor P. marinus was detected in this species until June, 3 mo after being transferred from the James River to the York River, making it likely that oysters were uninfected at the time of trans- fer. It is possible, however, that some early P. marinus infections were missed as a result of relying on histology for diagnosis rather than the standard thioglycoUate technique (Ray 1966) or a more sensitive hemolymph assay (Gauthierand Fisher 1990). From June through August, both the prevalence and the intensity of infections increased. In addition to high levels of H. nelsoni and P. marinus. the trematode B. cuculus was seen in 2 oysters, and 1 1 oysters were infected by more than one parasite species. In contrast, no parasites were seen in C. gigas at any time during this study. It is possible, however, that C. gigas did become infected with P. marinus. as noted previously (Meyers et al. 1991, Barber and Mann 1994), but infections never developed to the point where they were detectable by histological examination. C. virginica was thus susceptible and lacked resistance to all three parasite species, whereas C. gigas was either not susceptible or was highly resistant to the parasites. The lack of gametogenesis exhibited by C. virginica may in part be related to the source of the oysters used in this study. As described by Cox and Mann ( 1992), oysters on Horsehead Reef in the James River, VA, tend to have a lower fecundity compared with oysters inhabiting locations further downriver, probably the result of salinity, which averages less than 15 ppt. In spite of producing fewer eggs than oysters from other areas, however, Horsehead oysters showed evidence of spawning in both July and August (Cox and Mann 1992). In addition, there is no reason to suggest that oysters from Horsehead Reef would not be capable of undergoing a normal gametogenic cycle after transfer to the York River (salinity >20 ppt). A more likely explanation for the failure of C. virginica to undergo a complete gametogenic cycle is the combined negative effects of H. nelsoni and P. marinus and, to a lesser extent, B. cuculus, on reproduction. Several studies have noted a quantitative difference in gamete production between infected and uninfected individuals. For example. Barber et al. (1988a) found that com- pared with uninfected individuals, gamete production in oysters with epithelial and systemic H. nelsoni infections was reduced by 35 and 81%, respectively. Similarly, oysters with heavy P. mari- nus infections had significantly smaller gonadal areas than did uninfected oysters (Dittman 1993). Choi et al. (1994) found a negative correlation between the intensity of P. marinus infection and the rate of gonadal production. Although not considered pathogenic, sporocysts of the trematode B. cuculus typically in- vade oyster gonad tissue, causing castration (Menzel and Hopkins 1955, Sindenmann 1990). Both oysters containing sporocysts of B. cuculus in this study were devoid of gametes. Ahhough few C. virginica were free of parasites during the months of July through September, when the greatest gametogenic activity should have occured, differences between infected and uninfected individuals were noted. For example, the mean GAI of uninfected individuals was significantly greater (t-test, P S 0.01) than that of parasitized individuals from June through September. Further, in July, Au- Gametogenesis of C. virginica and C. gigas 289 gust, and September, when parasite prevalence and intensity were greatest. 14. 12. and 14 individuals, respectively, were cither in- active or actively resorbing gametes. This, in combination with the fact that parasites were not detected in the only mature individuals encountered in this study, strongly suggests that in the York River, most if not all gamete production in C. ivrx"'"" was overcome by the negative effects of both H. nelsoiii and P. nuirinus. Parasites might also have an effect on the timing of gameto- genic events of oysters. Ford and Figueras (1988) and Ford el al. (1990) found that in Delaware Bay. gametogenesis in C. virginica was inhibited in late spring when H. nelsimi levels were high, but as temperature-associated remission occurred in August and Sep- tember, many oysters developed mature gonads and spawned. In this study, however, only 5 (out of 45) oysters were undergoing gamete development from July to September. Although it is pos- sible that rapid development and spawning occurred between sam- ples and were undetected, it is more likely, given the large number of resorbing and undifferentiated individuals seen during this pe- riod, that oysters generally succumbed to the effects of parasites, which maintained a high intensity until October, at which time the temperature had dropped to below I5°C and was too low to sup- port gamete development and spawning. Thus, the combined ef- fects of W. nelsoni and P. mariniis particularly, both in terms of intensity and duration throughout the year, limited the ability of C. virginica to produce gametes and spawn. Observed differences in gametogenesis between C . virginica and C. gigas, especially those pertaining to the timing of events and rates of development, may have a genetic basis. It is known, for example, that there are genetic differences in the timing of gonadal maturation and spawning between populations of C. vi- rginica having different geographic origins (Barber et al. 1991. Ford et al. 1990); As seen in this study. C. gigas spawned once between mid-June and mid-July. According to Andrews (1979), C. virginica in Chesapeake Bay typically spawns over a 3-mo period (June to August). Differences in the quantitative aspects of gametogenesis such as fecundity and egg size might also differ between species. In this study, the mean GAI of mature C . vir- ginica was I37f . Much greater maximum mean GAIs of 30% and 35-45% were found by Barber et al. (1988a) and Barber et al. ( 1991 ). respectively, who compared gonadal development among several groups of C. virginica in Delaware Bay. It is unclear to what extent the difference between maximal GAI found here and those seen in previous studies is related to local environmental conditions versus levels of parasitism. In contrast, the fecundity of C. gigas appears to be much greater than that of C. virginica. The GAI of 43% found in this study for mature C. gigas. although much greater than that found for mature C. virginica, is less than the maximal values of 65-79% reported for C. gigas in previous studies (Mori 1979. Perdue et al. 1981 . Allen and Downing 1986), perhaps the result of environmental variations between locations. Another difference between species noted in this study was mean oocyte size. Mean oocyte areas of mature females were 1 ,236 (xm" for C. gigas and 820 |xm~ for C. virginica, which correspond to mean diameters of 39.7 and 32.3 jjim, respectively. Barber et al. (1988a) reported a maximum oocyte diameter for C. virginica from Delaware Bay of 36-37 ^jim. Even with a greater mean oocyte area, the larger GAI of C. gigas gives it a distinct advan- tage in terms of number of eggs produced. For mature oysters having a total cross-sectional area of 100 mm", 34,790 eggs would be bisected in the gonadal area of C. gigas compared with 15,850 eggs for C. virginica. Thus, in Chesapeake Bay waters, there are differences between oyster species in the timing of maturation and spawning as well as in relative fecundity, egg size, and egg num- ber, which are unrelated to the differential effects of parasitism. Given the fact that both H. nelsoni and P. mariniis have been endemic in lower Chesapeake Bay since 1959 (Andrews and Wood 1967; Andrews 1988) with no sign of abating, C. gigas appears to have a distinct advantage over C. virginica in terms of disease resistance (Meyers et al. 1 99 1; Barber and Mann 1994), growth (Barber and Mann 1994). and as seen in this study, repro- ductive potential. The production of C. virginica in Chesapeake Bay is unlikely to increase until both parasites diminish in viru- lence or resistance to both diseases is developed by the oyster host. ACKNOWLEDGMENTS Thanks to J. Walker and L. Ragone-Calvo (VIMS) for histo- logical processing and K. Walker (VIMS) for flume maintenance. This article is Maine Agricultural and Forestry Experiment Station external publication #1998 and contribution no. 1998 from VIMS. LITERATURE CITED Allen. S. K.. Jr. & S. L Downing. 1986. Pertbrmance of tnploid Pacific oysters, Crassostrea gigas (Thunberg). 1. Survival, growth, glycogen content, and sexual maturation in yearlings. J. Exp. Mar. Biol. Ecol. 102:197-208. Amemiya. I. 1928. Ecological studies of Japanese oysters, with a special reference to the salinity of Iheir habitats. J. Coll. Agric. Imperial Univ. Tokyo IX:333-381. Andrews. J. D. 1979. Pelecypoda: Ostreidae. pp. 293-341. In: A. C. Giese and J. S. Pearse (Eds.). Reproduction of Marine Imenehrates. Vol. V. Academic Press. New York. Andrews. J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marinus and its effect on Ihe oyster industry. Am. Fish. Soc. Spec. Piibl. 18:47-63. Andrews. J. D. & J. L. Wood. 1967. Oyster mortality studies in Virginia. VI. History and distribution of Minchinia nelsoni. a pathogen of oys- ters, in Virginia. Ches. Sci. 8:1-13. Barber. B. J. & N. J. Blake. 1983. Growth and reproduction of the bay scallop, Argopecten irradians (Lamarck) at its southern distributional limit. J. E.xp. Mar. Biol. Ecol. 66:247-256. Barber. B. J. & R. Mann. 1994. Growth and mortality of eastern oysters, Crassostrea virginica {Gmelin. 1791). and Pacific oysters, Crassos- trea gigas (Thunberg. 1793) under challenge from the parasite, Per- kinsus marinus. J. Shellfish Res. 13:109-114. Barber. B. J., S. E. Ford & H. H. Haskin. 1988a. Effects of the parasite MSX {Haplosporidium nelsoni) on oyster {Crassostrea virginica) en- ergy metabolism. I. Condition index and relative fecundity. J. Shellfish Res. 7:25-31. Barber. B. J.. R. Getchell, S. E. Shunway & D. Schick. 1988b. Reduced fecundity in a deep-water population of the giant scallop Placopeclen magellanicus in the Gulf of Maine. USA. Mar. Ecol. Prog. Ser. 42: 207-212. Barber. B. J., S. E. Ford & R. N. Wargo. 1991. Genetic variation in the timing of gonadal maturation and spawning of the eastern oyster, Cras- sostrea virginica (Gmelin). Biol. Bull. 181:216-221. Burreson. E. M., R. Mann & S. K. Allen. Jr. 1994. Field exposure of tnploid Crassostrea gigas to Haplosporidium nelsoni (MSX) and Per- kinsus marinus (Dermo) in the lower Chesapeake Bay. /. Shellfish Res. 13:293. Choi, K., E. N. Powell, D. H. Lewis & S. M. Ray. 1994. Instantaneous 290 Barber reproductive effort in female American oysters, Crassoslrea virginica. measured by a new immunoprecipitation assay. Biol. Bull. 186:41-61. Cox, C. & R. Mann. 1992. Temporal and spatial changes in fecundity of Eastern oysters, Crassoslrea virginica (Gmelin. 1791) in the James River, Virginia. J. Shellfish Res. 11:49-54. Dittman, D. E. 1993. The quantitative effects of Perkinsus marinus on reproduction and condition in the eastern oyster. Crassostrea virginica. J. Shellfish Res. 12:127. Ford, S. E. & A. J. Figueras. 1988. Effects of sublethal infection by the parasite Haplosporidium nelsoni (MSX) on gametogenesis, spawning, and sex ratios of oysters in Delaware Bay, USA. Dis. Aquat. Org. 4:121-133. Ford, S. E. & H. H. Haskin. 1982. History and epizootiology of Haplosporidium nelsoni (MSX), an oyster pathogen in Delaware Bay, 1957-1980. J. Inverlebr. Palhol. 40:118-141. Ford. S. £., A. J. Figueras & H. H. Haskin. 1990. Influence of selective breeding, geographic origin, and disease on gametogenesis and sex ratios of oysters, Crassostrea virginica. exposed to the parasite Haplosporidium nelsoni (MSX). Aquaculture 87:285-301. Gauthier, J. D. & W. S. Fisher. 1990. Hemolymph assay for diagnosis of Perkinsus marinus in oysters Crassoslrea virginica (Gmelin 1791). J. Shellfish Res. 9:367-371. Haskin, H. H. & J. D. Andrews. 1988. Uncertainties and speculations about the life cycle of the Eastern oyster pathogen Haplosporidium nelsoni (MSX). Am. Fish. Soc. Spec. Publ. 18:5-22. Haskin, H. H. & S. E. Ford. 1979. Development of resistance to Min- chinia nelsoni (MSX) mortality in laboratory-reared and native oyster stocks in Delaware Bay. Mar. Fish. Rev. 41:54-63. Haskin. H. H.,L. A.Stauber&J. A. Mackin. \966. Minchinia nelsoni n. sp. (Haplosporida. Haplosporidiidae): causative agent of the Delaware Bay oyster epizootic. Science 153:1414-1415. Howard. D. W. &C. S. Smith. 1983. Histological Techniques for Marine Bivalve Mollusks. NOAA Technical Memorandum NMFS-F/NEC-25, Oxford. Maryland. 97 pp. Kennedy, V. S. & H. I. Battle. 1964. Cyclic changes in the gonad of the American oyster, Crassostrea virginica (Gmelin). Can J. Zool. 42: 305-32 1 . Mann. R., E. M. Burreson & P. K. Baker. 1991. The decline of the Virginia oyster fishery in Chesapeake Bay: Considerations for intro- duction of a non-endemic species, Crassostrea gigas (Thunberg. 1793). J. Shellfish Res. 10:379-388. Menzel. R. W. & S. W. Hopkins. 1955. The growth of oysters parasitized by the fungus Dermocystidium marinum and by the trematode Buce- phalus cuculus. J. Parasilol. 41:333-342. Meyers, J. A., E. M. Burreson, B. J. Barber & R. Mann. 1991. Suscep- tibility of diploid and triploid Pacific oysters, Crassostrea gigas. to Perkinsus marinus. J. Shellfish Res. 10:433-437. Mon. K. 1979. Effects of artificial eutrophication on the metabolism of the Japanese oyster Crassostrea gigas. Mar. Biol. 53:361-369. Perdue. J. A., J. H. Beattie & K. K. Chew. 1981. Some relationships between gametogenic cycle and summer mortality phenomenon in the Pacific oyster {Crassostrea gigas) in Washington State. J. Shellfish Res. 1:9-16. Ray. S. M. 1966. A review of the culture method for detecting Dermo- cvstidiiini marinum. with suggested modifications and precautions. Proc. Natl. Shellfish Assoc. 54:55-69. Sindermann. C. J. 1990. Principal Diseases of Manne Fish and Shellfish Vol 2. Academic Press, San Diego, California. 516 pp. Wilkinson. L., M. Hill, J. P. Welna & G. K. Birkenbeuel. 1992. SYSTAT for Windows: Statistics, Version 5 Edition. SYSTAT. Inc., Evanston, Illinois. 750 pp. Journal of Shellfish Research. Vol. 15, No. 2. 291-295. 1996. THE SUITABILITY OF LAND-BASED EVALUATIONS OF CRASSOSTREA GIGAS (THUNBERG, 1793) AS AN INDICATOR OF PERFORMANCE IN THE FIELD GREGORY A. DEBROSSE AND STANDISH K. ALLEN, JR. Rutgers-the State Universit}- of New Jersey Institute of Marine and Coastal Sciences Haskin Shellfish Research Lab Box B-8. RD #1 Port Norris, New Jersey 08349 ABSTRACT The introduction of Crassosirea gigas to the mid-Atlantic requires prior knowledge of their likely ecological response, according to the International Council for Exploration of the Seas Committee guidelines. Without at least an experimental introduction, however, such knowledge is unattainable. Are comparisons of survival, growth, disease resistance, etc., conducted in land-based tanks suitable for estimating the performance of C. gigas in the field' In June 1991, equal numbers of spat from three crosses — MSX- resistant Crassosirea virginica (eastern), C. gigas form Miyagi. and C. gigas form Hiroshima — were split into two replicates and reared in upwellers for the first summer and in a land-based tank for the second. After the first season, C. virginica had the highest mortality (65. 36, and 13% for eastern, Miyagi, and Hiroshima, respectively) and average spat size was about 30% greater in both C. gigas groups. For the second year, the three crosses were transferred to a 16,000-L tank; two replicates of eastern oyster were also placed in Delaware Bay. Cumulative mortality for the second season (through 11/92) was eastern, 60%; Miyagi, 73%; Hiroshima, 93%; and eastern in Delaware Bay, 37%, In the tank, Miyagi oysters grew fastest, followed by Hiroshima and eastern; however, eastern oysters grown in the field were larger than all tank-reared groups. All oysters in the tank were infested with Polydora websleri, C ■ gigas heavily and eastern oysters lightly; eastern oysters grown in the field were virtually free of infestation. These data indicate that tank-based comparisons are unlikely to yield a true measure of performance in the local environment. KEY WORDS: Non-native species. Pacific oyster, eastern oyster, disease resistance, Miyagi, Hiroshima INTRODUCTION Crassosirea gigas (Thunberg. 1793). an oyster native to Japan but widely introduced around the world, has recently generated interest in the mid- Atlantic as a candidate for revitalizing the oys- ter fisheries of both Delaware and Chesapeake Bays (Mann et al. 1991. Lipton et al. 1992. Gaffney and Allen 1992). There are two principal ways that C. gigas might help. The first is by the direct introduction of the species as a replacement or supplemental fish- ery. This approach engenders the most controversy because it is (probably) irreversible using normal diploid populations. The sec- ond is the incorporation of useful genes from C. gigas. especially for disease resistance, into the native eastern oyster through hy- bridization or gene insertion. The idea that gene incorporation is a reasonable approach is indicated by recent evidence of disease resistance in C. gigas. C. gigas is more tolerant to Perkin.sus marimts (Dermo) than Crassosirea virginica (Meyers et al. 1991) and also seems to be resistant to Haplosporidium nelsoni (MSX) (Allen unpublished data, Burreson et al. 1994). So far, however, hybrids between eastern and Pacific oysters have failed (Allen et al. 1993). Whether direct introduction or gene introduction takes place. the field evaluation of pure Pacific oysters. Pacific-eastern hy- brids, or genetically modified eastern oysters will be essential. In addition to the question of disease resistance, there are a host of ecological questions regarding the suitability of the new candidate oyster for culture in the mid-Atlantic. Some have suggested that these tests need to be conducted in confined systems, i.e.. land- based and quarantined. This was especially true when proposals to test certified triploid Pacific oysters in the mid-Atlantic were ten- dered (Allen 1993). Would the same be required of hybrids or genetically modified oysters? If, for example, hybrids were pro- duced, would tests run in land-based tanks yield meaningful data on the likely field performance of candidate oysters? The follow- ing study was conducted to answer this question. MATERIALS AND METHODS 1991 Season In late May and early June 1991. larvae from three crosses were placed into culture at Rutgers Cape Shore hatchery: eastern oysters (C virginica). a cross between two Rutgers MSX-resistant lines; C. gigas form Miyagi. an F2 Washington state import; and C. gigas form Hiroshima, an F2 import from southern Japan. Larvae were reared in 211-L polypropylene tanks filled with l-|xm-pore-size filtered seawater at ambient temperature and sa- linity. During the larval culture period, seawater temperature ranged from 23 to 28°C and salinity ranged from 20 to 24 ppt. Larvae were fed a mixture of Isochrysis aff. galiiana. Thalassio- sira pseudonana. and Chaeloceros calcilrans at densities recom- mended by Breese and Malouf ( 1975). Tanks were drained every 48 h, at which time larval growth and survival rates were deter- mined. Miyagi had the highest survival to 48 h. but all groups had similar survival to the eyed stage (Table I ). All groups were set as cultchless oysters with epinephrine, generally following the guide- lines recommended by Coon et al. ( 1986). Recently set spat were held for 7-10 days in hatchery downwellers before rotation to a recirculating, upweller nursery system. Approximately 1 wk after rotation to the upweller. 4.130 animals from each of the three groups were taken from the size class retained by a 1.98-mm screen. Mean shell length (N = 50) was 3.8. 3.2. and 3.4 mm for eastern, Miyagi, and Hiroshima, respectively. The samples were split into two groups of 2,065. and each group was placed into a 30-cm-diameter x 46-cm-tall upweller silo receiving ~4 L/min of Delaware Day seawater, filtered to 100 (im. The relative position 291 292 DeBrosse and Allen TABLE 1. Spawning and larval survival data for eastern, Miyagi, and Hiroshima progeny groups used for this study. No. of Spaw lers No. of Eggs (millions) Percent Survival to Parent Stock Spawn Date Female Male 48 h Eyed Progeny Code DF X BLA 5/28/91 18 12 13 23 9 eastern XWAA 5/29/91 5 8 17 55 21 Miyagi ASJPNA 6/4/91 8 15 25 23 10 Hiroshima DF and BLA are seventh- and fifth-generation lines of MSX disease-resistant strains bred at the Haskin Shellfish Laboratory. XWAA is the second generation of C gigas form Miyagi imported from Washington in 1989. ASJPNA is a first-generation line of C. gigas form Hiroshima imported from Japan m 1990. of each silo within the raceway was rotated on a daily basis to minimize position effects. In July and August, upweller silos were graded every 2 wlc by the use of a sieve series ( 1 1 sieves; 1.52, 1.98. 2.87. 3.5, 5.54. 7.9, 8.98, 10.9, 12.0, 13.42, and 18.9 mm), at which time the total number of oysters and the total volume were calculated for each size class. In September and October, silos were graded every 4 wk. Incoming seawater tem- perature during the first season study period reached a maximum of 35. 1°C in late July. In late November, groups were enclosed in 2-mm plastic mesh bags and enclosed in wire mesh trays. All test groups were overwintered in a designated broodstock sanctuary. At the sanctuary site, seawater temperatures, ranging from a low of 1. 0°C (12/91) to a high of 4. IT (3/92). were recorded during weekly sampling periods. Salinity was approximately 30 ppt. 1992 Season In April 1992. the groups were transferred to a 1.5-m-tall x 3.7-m-diameter, 16,000-L holding tank. Oysters were held in 104- cm-long X 48-cm-wide x 11.5-cm-deep wire mesh trays lined with 6-mm plastic mesh. The position of the trays within the tank was rotated on a daily basis to minimize position effects. The tank was drained into an underground line, cleaned, and refilled with raw seawater daily. In late May, after a large-scale mortality event, a 5-cm airline was added as a source of aeration. From April through November, live counts were taken every 2 wk; the shell length (to the nearest 0.1 mm) and whole weight (to the nearest 0.1 g) of 25 oysters in each replicate were measured monthly. This regime was also followed for two replicates of eastern oysters grown in wire mesh trays (as described above) on adjacent tidal flats in Delaware Bay. Beginning in early September, we noticed that the interior shell surface of all C. gigas mortalities was covered with vesicles cre- ated by the mud worm Polydora websterl. On November 9, 1992, 25 live animals from each of ten groups (N = 250) were sacrificed to quantify P. websteri incidence. Oysters were opened, and each valve was scored as having light incidence (presence of isolated blisters only), heavy incidence (multilayer continuous vesicles covering the entire shell surface), or no incidence. Percent inci- dence was calculated by dividing the number of oysters having light (or heavy) infections by the total number of oysters examined for each group. The maximum incoming seawater temperature during the second season study period was 30.5°C (reached on ,M..; 25 and again on July 16). Statistical Analyses All data were analyzed with the computer program SYSTAT (Wilkinson 1990) by analysis of variance (ANOVA). followed by Tukey's HSD Multiple Comparisons. Survival data were arcsine transformed before statistical analysis (Sokal and Rohlf 1981). RESULTS 1991 Season Mortality Most of the mortality for all groups occurred during an approx- imately 30-d period from Day 15 to Day 42 (Fig. 1 ). At the end of this period. Hiroshima had significantly higher survival than east- em or Miyagi oysters (Tukey"s HSD = 0.033). After 98 d, cu- mulative percent mortality totaled 65, 36, and 13% for eastern, Miyagi, and Hiroshima, respectively. Cumulative mortality in both C. gigas races was significantly lower than in the C. virginica cross (Tukey's HSD, p =0.005 with Hiroshima, 0.021 with Mi- yagi). Growth Even though all oyster spat were initially taken from spat re- tained on a 1.98-mm sieve, the initial starting volume was 30% m o E E o uu - Eastern : Mi /agl 80 - Hiroshima • • 60 - • ""^ • 40 - T ■ ■ ■ r* ■ ■ 20 - myj ♦ X ♦ - > 0 -. 1^-^* 1 1 1 Jul Aug Sep 1991 Oct Nov Figure \. Cumulative percent mortality (symbols) of spat by replica- tion for eastern (C. virginica) (circles) and C. gigas, forms Miyagi (squares) and Hiroshima (diamonds), from July 9 (Day 0, first de- ployed as spat) to October 15, 1991 (Day 98), in upweller silos. Mean cumulative percent mortality is depicted by the lines. Performance of C. gigas in Tanks 293 less for Miyagi (38 ml) and 19% less for Hiroshima (44 mil com- pared with eastern (54.4 ml). This was the result, as mentioned previously, of the somewhat smaller initial mean spat size for Miyagi (3.2 mm) and Hiroshima (3.4 mm) than that for eastern (3.8 mm). The subsequent increase in total volume in the two C. gigas groups exceeded that of C. virgtnica during the first season. At Day 98. total pooled volume was approximately 3.2 times greater in Hiroshima (5.265 ml) and 2.5 times greater in Miyagi (4.134 ml) compared with eastern (1.623 ml). This comparison, however, does not take into account that by Day 98. there were about twice as many spat left in the C. gigas groups as in the C. virginica replicates. We standardized the increases in the volume of the spat by dividing the total volume by the total number of spat, yieldmg a measure of average spat volume. The mcrease in spat volume of both C. gigas groups exceeded that of C . virginica (Fig. 2), despite the twofold difference in density in the upweller silos. ANOVA showed a significant difference among the groups for average spat volume, but a Tukey's test failed to show a significant difference between the C. gigas races and eastern oys- ters (probabilities for pairwise comparisons: Miyagi vs. eastern, p = 0.055; Hiroshima vs. eastern, p = 0.064). A breakdown of oyster populations by size class showed that 86% of eastern. 98% of Miyagi, and 97% of Hiroshima spat were larger than 9 mm by mid-November (Fig. 3). 3 cr 0) 60 50 40 30 20 10 Eastern Miyagi Hiroshima Screen size (mm) Figure 3. Size class distributiuns for eastern (C. virginica) and C. gigas, forms Miyagi and Hiroshima, on October 15, 1991, 98 days after deployment into upweller silos. Second Season Mortality A large-scale mortality event began during the third week of May and continued through June (Fig. 4). Mortality was highest in Hiroshima oysters at nearly every sampling period, significantly higher than both eastern groups in June. July, and August. The survival of Miyagi oysters was intermediate between Hiroshima and tank-held eastern. Cumulative percent mortality from May through June was 43% for eastern, 56% for Miyagi. and 83% for r 3 CO a. m Q. 0) E _2 o > O) (D < 1 - 0 - Eastern Miyagi Hiroshima Jul Aug Sep 1991 Oct Nov Figure 2. Average spat size (total volume/total number of spat) for eastern (C. virginica) and C. gigas, forms Miyagi and Hiroshima, on October 15, 1991, 98 d after deployment into upweller silos. Hiroshima, accounting for the majority of mortality within tanks. The corresponding mortality for the same period of time in the replicates of eastern oysters grown in Delaware Bay was 3%. The cumulative percent mortality from April 1992 to November 1992 was 60% for eastern. 73% for Miyagi. 937f for Hiroshima, and 37% for easterns held in Delaware Bay. Growth During the second season. Miyagi grew fastest, followed by Hiroshima and eastern; however, easterns grown on the tidal flats were larger than all tank-reared groups (Fig. 5). Mean shell length (in millimeters) and mean whole weight (in grams) at the end of the study period ( 1 1/92) was 44.7 mm and 17.6 g for eastern. 51.9 mm and 22.4 g for Miyagi. 45.7 mm and 16.8 g for Hiroshima, and 56.0 mm and 28.2 g for eastern oysters grown in Delaware Bay. The weight of field-grown easterns was significantly larger than that of the other groups in September. October, and Novem- ber. P. websteri Incidence On November 9, 1992. it was determined that multilayer ves- icles covered the entire shell surface of the flat valve in 87% of Miyagi and 94% of Hiroshima oysters (Fig. 6). No vesicles were found in either eastern group. However, isolated blisters were found on the flat valve of 66% of tank-reared eastern and in only 2% of those animals grown in Delaware Bay. DISCUSSION The objective of this study was to evaluate whether land-based comparisons of growth and survival in tanks are reasonable esti- mations of these same measures in the field. Obviously, because the C. gigas races could not be deployed in the field, there was no corresponding field group for this species. We can only extrapo- late on how C. gigas would have performed in the field on the basis of the comparison of tank- and field-grown eastern oysters. 294 DeBrosse and Allen ■e o E 60 - 40 20 Eastern (tank) • Miyagi ■ Hiroshima ♦ Eastern (field) o I I Light Heavy — 1 1 1 1 1 \ \ < Apr May Jun Jul Aug Sep Oct Nov 1992 Figure 4. Percent mortality by month (symbols, by replication) from April 29 to November 9, 1992, for eastern (C. virginica) and C. gigas. forms Miyagi and Hiroshima, juveniles grown in tanks and for eastern oysters grown in the field in Delaware Bay [Eastern (field)]. Mean percent mortality is depicted by the lines. This study also provided an opportunity to compare the pert'or- mance of C. virginica vs. two races of C. gigas "head to head" in conditions mimicking Delaware Bay, albeit in tanks only. The performance of larvae in the hatchery was similar among the three groups, although cultures were not replicated and statis- tical comparisons could not be run. The survival of C. gigas larvae to 48 h equaled (Hiroshima) or exceeded (Miyagi) that of C. I o Eastern (tank) • Miyagi ■ 30 - Hiroshima ♦ o Eastern (field) o 0 1 ■ 20 - t 1 ' o ■ 10 - a 1 V • y 0 - 1 1 1 1 1 1 1 1 Apr May Jun Jul Aug Sep 1992 Oct Nov Dec Figure 5. Whole weight (symbols), by replication, of eastern (C. vir- ginica) and C. gigas, forms Miyagi and Hiroshima, juveniles and east- ern oysters grown in the field in Delaware Bay [Eastern (field)] from April to November 1992. Mean whole weight is depicted by the lines. 100 80 60 40 - 20 E, M H Ef E, M H Ef Cupped valve Flat valve Figure 6. Mean percent incidence of P. websleri infections found in cupped and flat shells from eastern (C. virginica) (E,) and C. gigas, forms Miyagi (M) and Hiroshima (H), oysters grown in a tank and from eastern oysters grown in the field (E^). Isolated blisters were scored as light infections; vesicles covering the shell surface were scored as heavy infections. virginica. but survival to the eyed stage was similar in all groups. After setting, C. gigas were fast starters and seem to thrive in the warmer waters during juvenile periods. After 4 mo in the upweller nursery system, C. gigas spat survived better and average spat size was about i0'7c greater than the C. virginica group. It is relevant to note at this point that spat in our upweller system are normally rotated to our field grow-out system when they reach a size of approximately 9.0 mm. This normally occurs within 3-6 wk. Four weeks after rotation to the upweller system, about a quarter of C. virginica spat were ^9.0 mm. whereas about half of the C. gigas groups were of appropriate size. Holding all groups in this system for an additional 3 mo subjected them to a food-limiting environ- ment during that period. It is likely that, had we been able to rotate these spat to the field according to our normal grading routine, the size differences between the C. gigas and the C. virginica would have been even greater than those observed. The results of the second study period presented an entirely different outcome. Large-scale mortality began during the third week of May and continued through June, affecting particularly C. gigas, and accounted for the majonty of the mortality during the 2-yr period. We were surprised by this early season mortality for a number of reasons. First, study groups were housed in a 16,000-L tank that was drained and refilled every 24 h. There was little chance of food deprivation, and the tank was vigorously aerated beginning 5/22, an extraordinary precaution not used in other years. Second, these groups were stocked at approximately one-third the density at which we normally hold tank-reared groups, discounting the possibility that waste product build-up was problematic. Finally, there was no evidence of disease when nearly dead oysters were examined (S. Ford, HSRL, Rutgers Uni- versity). We can only speculate that the mortality event was due to the unseasonably warm temperatures during May 1992 (maximum incoming seawater temperature occurred on May 25). We further Performance of C. gigas in Tanks 295 speculate that the mortality was associated with gonad production in the Pacific oysters, a phenomenon referred to as ■'summer mortality" in commercially grown oysters on the West Coast of the United States (Perdue 1983). Whatever the reason, the out- come was to reverse the order of the group standings regarding cumulative percent mortality. Additionally, mortality was prefer- entially higher in the larger size classes of oysters (personal ob- servation), which greatly reduced the size differences between C. gigas and C. virginica observed before the mortality event. A final unexpected variable encountered during the second season was the extremely high incidence and seventy oi P . websteri infestation in oysters, especially C. gigas held in tanks. The higher incidence of Polydora in the tank-grown oysters was most likely the result of the recruitment of late-stage Polydora larvae brought into the tank system in the daily delivery of unfiltered seawater. Lunz ( 1941 ). in an article examining Polydora infestation in South Carolina oys- ters, indicated that a large number of mud blisters within the shell may restrict the living space of the oyster and that the animal may be forced to spend considerable energy in secreting shell material for covering the mud worms. Heavy infestations may also cause heavy mortality (Roughley 1922). Thus, it is reasonable to assume that the severe Polydora infestations found among C. gigas in particular had a deleterious effect on both survival and growth. Differences in growth were apparent among all three groups of oysters held in the tank. Growth and survival differed between Miyagi and Hiroshima oysters, supporting their status as physio- logical races. Compared with eastern oysters grown in the field, those in the tanks performed poorly for growth, survival, and even P . websteri infestation. Extrapolating from how well C. gigas did in the tank during the first season to how C. gigas might have done in the field might cause us to estimate that they would have done extraordinarily well. On the other hand, the presence of Polydora in the tank and not in the field and the extraordinary mortality of C. gigas in the tank during the second season of study signify that the two environments are not sufficiently similar to make such extrapolations. Although it may be possible to engineer a system that includes extensive filtration to remove abundant larvae and lowering the temperature of incoming seawater with chillers, at some point, the system no longer models "■field" conditions. We can only conclude that tank-based comparisons are not likely to generate a true estimate of growth and survival, and perhaps re- sponse to disease, in the local environment. LITERATURE CITED Allen. S. K.. Jr. 1993. Tnploids for field tests' The good, the bad. and the ugly. J. Shellfish Res. 12: i:.*! (abstract). Allen. S, K.. Jr.. P. M. Gaffney. J. Scarpa & D. Bushek. 1993. Inviable hybrids of Crassoslrea virginica (Gmelin) with C. rivuUiris (Gould) and C. gigas (Thunberg). Aquacidlure 113;269-289. Breese. W. P. & R, E. Malouf. 1973. Hatchery Manual for the Pacific Oyster. Oregon State U, Publ. ORESU-H-75-002. Corvallis. Oregon. 22 pp. Burreson, E., R. Mann & S. K. Allen. Jr. 1994. Field exposure of triploid Crassoslrea gigas to Haplosporidium nesloiri (MSX) and Perkinsus marinus (Demio) in the lower Chesapeake Bay. J. Shellfish Res. 13; 293 (abstract), Coon.S, L..D- B Bonar&R, M, Weiner, 1986. Chemical production of cultchless oyster spat using epinephnne and norepinephnne, Aquacul- ture 58:255-262, Gaffney, P. M. & S. K. Allen. Jr. 1992. Genetic aspects of introduction and transfer of molluscs. J. Shellfish Res. 1 1:535-538, Lipton. D. W,. E, F, Lavan & I, E, Strand. 1992, Economics of mollus- can introductions and transfers: the Chesapeake Bay dilemma. J . Shell- fish Res. 11:511-519. Lunz. G, R,. Jr, 1941, Polydora. a pest in South Carolina oysters. J. Elisha Mitchell Sci. Soc. 57:273-283, Mann. R,. E, M, Burreson & P, K, Baker. 1991. The decline of the Virginia oyster fishery in Chesapeake Bay: Considerations for intro- duction of a non-endemic species, Crassoslrea gigas (Thunberg. 1793), J. Shellfish Res. 10:379-388, Meyers, J. A., E. M. Burreson, B. J. Barber & R, Mann, 1991, Suscep- tibility of diploid and tnploid Pacific oysters, Crassoslrea gigas (Thun- berg. 1793) and eastern oysters, Crassoslrea virginica (Gmelin 1791), to Perkinsus marinus. J Shellfish Res. 10:433-437, Perdue, J. A, 1983, The relationship between the gametogenic cycle of the Pacific oyster. Crassoslrea gigas. and the summer mortality phenom- enon in strains of selectively bred oysters. Ph.D, Dissertation, Univer- sity of Washington. Seattle, Washington. 205 pp. Roughley, T.C. 1922, Oyster Culture on the George's River, New South Wales, pp. 1-69. Technical Education Senes, No. 25. Technological Museum. Sydney, Sokal. R, R. & F. J. Rohlf, 1981, Biometry, W, H, Freeman and Com- pany. New York, 859 pp, Wilkinson. L, 1990. SYSTAT: The System for Statistics. SYSTAT, Inc., Evanston. Illinois, 676 pp. Journal of Shellfish Research. Vol. 15, No. 2. 297-303. 1996. SPAWNING CYCLE OF THE PEARL OYSTER, PINCTADA MAZATLANICA (HANLEY, 1856), (PTERIIDAE) AT ISLA ESPIRITU SANTO, BAJA CALIFORNIA SUR, MEXICO FEDERICO GARCIA-DOMINGUEZ, BERTHA PATRICIA CEBALLOS-VAZQUEZ, AND ARTURO TRIPP QUEZADA Centra Interdisciplinario de Ciencias Marinas InstUuto Poliiecnico Nacional Apdo. Postal 592. La Paz. B.C.S. 23000. Mexico ABSTRACT The annual reproductive cycle of the pearl oyster, Pinctada mazatlanica (Hanleyl. from I.sia Espirilu Santo, B.C.S. , Mexico, was examined. Pearl oysters were collected at monthly intervals from June 1992 through August 1993. The gonadal development was analyzed by histological techniques and analysis of oocyte size. Spawning took place throughout the year, hut at a lower rale in winter. A relationship between spawning and temperature was not observed. The sex ratio was equal. KEY WORDS: Pearl oysters, spawning, Pinclada mazatlanua. reproduction INTRODUCTION The pearl oyster. Pmctada mazatlanua (Hanlcy, 1856), is a Panamic bivalve ranging from the outer coast of Baja California, through the Gulf of Cahfomia, and south to Peru (Keen 1971). In the last century, near La Paz in the Gulf of California, this species was abundant and supported a thriving pearl-fishing industry. The quality of the pearls produced by this species was a factor of great economic importance in the foundation and colonization of South Baja California (Cariho and Caceres-Marti'nez 1990, Monteforte and Carino 1992). The Mexican government closed the fishery in 1938 because of overfishing (Sevilla 1969. Keen 1971. Monte- forte 1990). P. mazatlanica is a potential mariculture species for the Gulf of California (Monteforte 1990). For resource management or mariculture purposes, it is impor- tant to know the life cycle of the target species. Documentation of the reproductive cycle of P. mazatlanica is crucial for a better understanding of the population dynamics that regulate the remain- ing wild stocks. Studies of reproduction by using histological tech- niques and light microscopy have been carried out on pearl oysters like Pinctada maxima (Rose et al. 1990), Pinctada fucata (Tranter 1959), Pinctada alhina (Tranter 1958a. Tranter 1958b, Tranter 1958c), and Pinctada margaritifera (Tranter 1958d). Previous to this, the only reproductive study in a natural population of P. mazatlanica was that of Sevilla (1969), who worked on a now extinct population located in La Paz harbor navigation channel, B.C.S. Mexico. For cultured pearl oysters {P. mazatlanica). two studies of reproduction were made in Bahia de La Paz (Saucedo and Monteforte 1994. Saucedo 1995). The purpose of this study was to determine the annual reproductive cycle of a wild adult population of P. mazatlanica at Isla Espiritu Santo, B.C.S., Mex- ico. MATERIALS AND METHODS From June 1992 to August 1993 (except in November 1992), 17-20 adult specimens per month of pearl oysters, P. mazatlanica, were collected randomly from a wild population located near Isla Espiritu Santo. B.C.S., (Fig. 1) by SCUBA from a 3- to 4-m depth. A total of 308 organisms were captured ranging from 64 to 161 mm in shell length (mean, 126.5: SD, 18.6) and from 72 to 176 mm in shell height (mean, 136.3; SD. 20.6). At the time of the collection of biological samples, water temperature was re- corded. Before dissection, shell height and shell length was measured with a 0.01-mm resolution caliper. The visceral mass (gonad in- cluded) was fixed in a neutral solution of 107f formalin prepared with seawater, dehydrated in an alcohol series dilution, and em- bedded in paraffin (Luna 1968). Sections 7 to 9 (j.m thick were made and stained with Harris' hematoxylin & eosin (Luna 1968). In addition, the diameter of at least 100 oocytes was measured, with an eyepiece graticule calibrated with a stage micrometer, in each of seven females per month selected randomly. The mea- surements were made along the longest axis in the oocytes sec- tioned through the nucleus containing clearly visible nucleoli. From these data, mean oocyte size and standard deviation were obtained. Individuals with few measurable oocytes and extensive phagocytosis ("spent" specimens) were not considered, following the criteria of Grant and Tyler (1983a. 1983b). Sex was determined by histological analysis. The percentage of each sex during the study period was obtained. The sex ratio was determined and was examined for deviation from the expected ratio of 1:1 by x~ analyses (Snedecor 1950). Categories of Gonadal Condition Gametogenesis (either spermatogenesis or oogenesis) of P. mazatlanica was divided into five stages (indifferent, developing, ripe, partially spawned, and spent) on the basis of the classifica- tion cited for P. mazatlanica (Sevilla 1969. Saucedo and Monte- forte 1994. Saucedo 1995). P. alhina (Tranter 1958a, Tranter 1958b), P. margaritifera (Tranter 1958d), P. fucata (Tranter 1959), and P. maxima (Rose et al. 1990). The relative frequency of the gonad developmental phase was determined. Developmental Stages Indifferent Stage. All of the oysters examined were adults (Fig. 2). In both male and female individuals, there was no evidence of gonadal development. It was not possible to distinguish the sex. The connective tissues occupied all of the space between the empty and collapsed follicles. 297 29S Garcta-Dominguez et al. ISLA ESPIRITU SANTO MEXICO Figure 1. Study area. Isia Espiritu Santo, B.C.S. marks the sampling area. Mexico. Asterisk lumen) increased, the amount of connective tissue decreased. The developing oocytes, which began as hemispherical stalked cells attached to the wall of the follicle, became enlarged spherical cells. 52.5 |j.m (SD. 4.7) in diameter, as maturity approached. Ripe Stage The ripe ovary was characterized by the presence of distended follicles filled with ripe poligonal-shaped oocytes, some of which were attached to the follicular wall by slender stalks (Fig. 4b). Little or no connective tissue was present between the follicular walls. Partially Spawned Stage Some follicles contained oocytes, whereas others were empty (Fig. 4c). There was a large reduction in the number of free large oocytes present in the lumen. Little connective tissue was present. Some follicular walls were broken. Spent Stage The follicles were empty, with the exception of a few large, unspent oocytes free in the lumen being phagocytized by amebo- cytes (Fig. 4d). Follicular walls were reestablished. RESULTS Developmental Stages of the Male Developing Stage The follicles, in different stages of development, occurred be- tween the connective tissue (Fig. 3a). Inside the follicle, a variable quantity of germinal cells and ripe gametes were observed. Sper- matozoa were stored as a dense mass in the lumen of the follicle, with the eosinophilic tails projecting into the central lumen. The area of connective tissue between the follicles decreases while follicles increase in area as the result of the accumulation of sperm. Ripe Stage The follicles were distended and filled with dense spermatozoa (Fig. 3b). Spermatocytes and spermatids were restricted to a thick layer on the follicular walls. Almost all connective tissue between follicles was replaced by follicles full of spermatozoa. Partially Spawned Stage During this reproductive stage (Fig. 3c), spermatozoa are ex- pelled into the environment. The follicles were partially empty. and their walls became broken. There was a marked decrease in the number of spermatozoa filling the lumen. Spent Stage The follicles are collapsed, have decreased in area, and are invaded by amebocytes (Fig. 3d) which phagocytize the unspent spermatozoa. There was no evidence of active spermatogenesis taking place. Developmental Stages of the Female Developing Stage Follicles were visible (Fig. 4a). Oocytes inside them increased in size and number. As the number of mature ova (free in the Reproductive Cycle The annual reproductive cycle of P. mazatlanUa is summa- rized in Figure 5. The gametogenic development of this species indicated spawning activity throughout the year. Partially spawned individuals were observed all year except in January, March, and July 1993. The highest frequency of partially spawned individuals was observed in June. July, October 1992, and April 1993, when the spawning individuals were 28.6, 33.3. 29.2. and 21.2% of the population and the temperature was 28. 29. 30. and 23°C. respec- tively. Pearl oysters in the indifferent stage were observed in July 1992 through January 1993 and in August 1993. Developing pearl oysters were found every month, except August 1992. Ripe oys- ters were encountered throughout the study except in August ^ / fWi X. t Figure 2. Indifferent stage. Scale bar = 50 p,m. Spawning Cycle of P. mazatlanica 299 i"*' it'^-z ^ ^-■ » Figure 3. Photomicrographs of gonadal stages of the male pearl oyster P. mazatlanica. (a) Developing stage, (b) Ripe stage, (c) Partiall.v spawned stage, (d) Spent stage. Scale bar = 50 \x.m. through December 1992. The largest number of ripe pearl oysters occurred in March 1993 (52.99'f of the individuals). The spent stage was recorded during June through December 1992 and May. July, and August 1993. Two hermaphroditic specimens were encountered. The micro- scopic examination revealed the presence of oocytes in a gonad with primarily spermatogenic development (Fig. 6a). In the other gonad, spermatozoa were found in the center of follicles of obvi- ous female development (Fig. 6b). Analysis of Oocyte Size lowed by a decrease in oocyte diameter, which indicates a spawn- ing event. Sex Ratio Females outnumbered males. A total of 308 pearl oysters were sampled, of which 159 (51.62%) were females. 119 (38.63%) were males, two were hermaphroditic (0.64%), and 28 (9.1%) were undifferentiated. The sex ratio ( 1 .33F: IM. n = 278) did not differ significantly (P 3= 0.01) from the expected ratio of l;l. Temperature The mean of oocyte diameter values for the study period can be During the study period, the water temperature varied from 21 observed in Figure 7. Maximum diameters were observed in June to 31°C. The highest values were in September 1992, and the 1992, April 1993, and June 1993. All of these peaks were fol- lowest values were in February 1993 (Fig. 8). 300 Garcia-Dominguez et al. %^J^. ^v^ 4S Figure 4. Photomicrographs of gonadal stages of the female pearl oyster P. mazatlanica. (a) Developing stage, (b) Ripe stage, (c) Partially spawned stage, (d) Spent stage. Scale bar = 50 |jim. DISCUSSION The characteristics of gametogenesis in P . mazatUinica were similar to those described for P. albina (Tranter 1958b), P . mar- garitifera (Tranter 1958c). P. fucata (Tranter 1959). and P. max- ima (Rose et al. 1990). During early gametogenic development, the follicles increased in length and volume as the number of mature ova or spermatozoa increased. As in P. albina (Tranter 1958a), in P. mazatlanica. sections across the gonad revealed that it consists of a system of branched tubules, the number and size of which depend on the stage of gonadal development. Older follicles contain fewer stem cells but many gametes. The pearl oysters P . maxima. P. albina. and P. margarilifera are protandric hermaphrodites (Tranter 1958c. Tranter 1958d, Rose et al. 1990). whereas, in P. fucata. both protandric and protogynic sex changes were recorded (Tranter 1959). P. mazat- lanica is also a protandric hermaphrodite (Sevilla 1969. Saucedo and Monteforte 1994). In a study with cultured young adults. oysters smaller than 100 mm matured as males, with females not occurring until animals attained a size larger than 100 mm. The sex ratio of cultured young adult pearl oysters was 0.12F;1M (Saucedo and Monteforte 1994). In this study, the sex ratio was 1.33F;1M. This work was carried out with larger individuals (mean height. 136.3 mm); however, females of 82-. 91-. 96-. and 98-mm shell height were found. The results of histological gonad examinations and oocytes diameter analysis indicate that P. mazatlanica from Isla Espiritu Santo, B.C.S., collected from June 1992 to August 1993 showed no clearly defined seasonal reproductive cycle. Spawning occurred throuahout the year, but on a minor scale in winter. Previous Spawning Cycle of P. mazatlanica 301 100 3 o UJ > I- < SPENT PART SPAWN RIPE DEVELOPING INDIFFERENT J J ASOD J FMAM J J A JUN 1992- AUG 1993 Figure 5. Relative frequency of gonadal stages of P. mazatlanica dur- ing June 1992 to August 1993. Observations of males and females were combined. researchers using histological analysis have reported that P . maza- tlanica spawn only in summer (Sevilla 1969, Saucedo and Mon- teforte 1994. Saucedo 1995). Those studies were performed on young individuals, predominantly males (77%). kept under con- ditions of experimental culture in bottom cages (Saucedo and Monteforte 1994. Saucedo 1995) or with only females collected in 1963 in the La Paz harbor navigation channel (Sevilla 1969). At this time, that population has disappeared (personal observation). Monteforte and Cariho (1992) also reported an absence of indi- viduals in La Paz Channel as well an impoverished number in other local populations of La Paz Bay. The histological observations of the spawning in P mazatlan- ica are supported by studies of spat settlement. P . mazatlanica settlement was reported from July to November in spat collectors with 5 and 8 wk of immersion (Caceres-Martinez et al. 1992. Monteforte and Garcia-Gasca 1994). June to October in collectors with 8-10 wk of immersion (Monteforte and Wright 1994). and in January and August to December in collectors with 8-10 wk of immersion (Felix-Pico 1977). The age of the largest pearl oysters from a collector is approximately equal to its period of exposure (Tranter 1958a). The period of free-swimming larval existence in P. mazatlanica is about 3 wk (Mason-Suastegui 1987). In P . maxima, mature individuals were observed outside the main breeding period, during the cooler months (Wada 1953, quoted in Rose et al. 1990). Similarly, in pearl oysters from Isla Espiritu Santo, ripe gonads were observed in winter, also outside the main breeding period. Histological data suggest that cultured P. mazatlanica (Saucedo 1995) were capable of spawning throughout the year. Sevilla (1969) indicates that the spawning temperature of P. mazatlanica is 27-29°C, with a maximum spawning at 28-29°C. This author also reported that spawnmg ends when the temperature is less than 25°C. Saucedo and Monteforte (1994) reported that cultured P. mazatlanica spawned when water temperatures reached 28-30°C. We observed spawning at temperatures ranging from 21 to 3rC. Temperature is one important environmental factor in the reg- ulation of bivalve reproduction (Sastry 1979). The influence of temperature on the reproductive cycles of other bivalves from Baja California Sur, Mexico, has been well documented (Garcia- Dominguez et al. 1993, Garcia-Dominguez et al. 1994. Villalejo- ^^K<^^!Atjl?# Figure 6. Hermaphroditic specimens, (a) Testis in gametogenesis with sperm surrounding an oocyte (arrow), (b) Ovary in gametogenesis with residual spermatozoa (arrows) occupying center of follicle. Scale bar = 25 jxm. Fuerte and Ochoa-Baez 1993, Villalejo-Fuerte et al. 1995). In this work, we did not observe a clear relationship between temperature and spawning. Similarly, Garcia-Dominguez et al. (1994) did not observe a clear relation between temperature and spawning in Megapilaria aurantiaca. also from Isla Espirutu Santo. The importance of food availability has been emphasized in the timing of bivalve reproduction (Bayne and Newell 1983, Mac- Donald and Thompson 1985. Jaramillo et al. 1993). Time of spawning may be related to food availability (Jaramillo et al. 1993). The spawning in bivalves might be synchronized to coin- cide with maximum food availability for larval development (Seed 1976. Jaramillo et al. 1993). In Chlamys amandi. the spawning time appeared to be related to high food levels rather than to water temperature (Jaramillo et al. 1993). whereas in Hinnites gigan- teus. the histological data suggest that there is no correlation be- tween food availability and spawning (Malachowski 1988). In P. mazatlanica. the maximum spawning time (summer) did not co- incide with maximum winter food availability. Signoret and San- toyo ( 1980) reported for La Paz Bay a maximum abundance of phytoplankton in winter (1,708,950 cells/1) and a minimum in spring (140,000 cells/I). Tranter (1958d) reported the spawning period for P. margari- 302 Garcia-Dominguez et al. 60 CO LJJ N CO 40 LJJ \- > O 30 o o 20 1UUi — 90- ■^\^ 80- ,^ \ \ * ss 70 ^■- .^ 1 1 - CO 60 \ 1 7 \ 1 ■Z bO \ 1 40- \ \ Q. CO 30- 20- 10- 0 m' -A V X X V ,■-■' K ^^ J J A S 0 D J F M A M J J A JUN1992 AUG 1 993 - --- SPAWNING ~ TEMPERATURE 32 30 O J J A ASODJ FMAMJ J JUN 1992 -AUG 1993 Figure 7. Mean oocyte sizes of P. mazallanica between June 1992 and August 1993. Bar = standard deviation. tifera in summer and winter, and in intermediate periods at a reduced intensity. Tiie similarity of their results with those ob- tained in this study for P. mazatlanica was most likely the result of the genetic closeness of the two species (Jabbour 1988). The breeding seasons of other species of Pincuida differ — P. maxima breeds annually, with maximal and minimal developmental peri- ods consistent with correspondingly high and low temperatures (Rose et al. 1990); P. albina breeds continuously throughout the year, but most actively in April and May (autumn) when sea 28 26 24 22 20 Figure 8. Relation of partially spawned stage with water temperature. Observations of males and females were combined. temperature begins to fall (Tranter 1958b); and P.fucata breeds in summer and autumn (Tranter 1959). ACKNOWLEDGMENTS Our gratitude to the Direccion de Estudios de Posgrado e In- vestigacion del Institute Politecnico Nacional (IPN). who provided funds for this woric. to Jose Luis Castro Ortiz for his help collect- ing samples, and to Consuelo Gonzalez O. for her editorial help with the English manuscript. We acknowledge the fellowships of Comision de Operacion y Fomento de Actividades Academicas to F. Garcia-Dominguez and A. Tripp-Quezada and Programa Insti- tucional de Formacion de Investigadores to B. P. Ceballos- Vazquez. Thanks to anonymous reviewers for their constructive comments on the manuscript. LITERATURE CITED Bayne. B. L. & R. C. Newell. 1983. Physiological energetics of marine molluscs, pp. 491^98. In: A. S. M. Saleuddin and K. M. Wilbur (eds.). The Mollusca. vol. 4. Academic Press, New York. Caceres-Martinez, C. C. Ruiz-Verduzco & D. Ramfrez-Filippini. 1992. Experimental collection of pearl oyster. Pmctada mazatlanica and Pte- ria sterna, spat on a filament substrate J. World Aquacultiire Soc. 23(3):232-240. Cariiio. M. & C. Caceres-Martinez. 1990. La Perlicultura en la Peninsula de Baja California a principios de siglo. Serie Cienlifica U.A.B.C.S. (Niimero Especial AMAC) 1:1-6. Felix-Pico, E. F. 1977. Informe final del proyecto de cultivo de moluscos (Bahia de La Paz). Informe Tecnico. Subs. Pesca. Mexico (Ed.) Depto. de Pesca. La Paz. B.C.S. 48 pp. Garcia-Dominguez, F., G. Garcia-Melgar & P. Gonzalez-Ramirez. 1993. Reproductive cycle of the clam Chione californiensis (Broderip, 1835), in Bahia Magdalena, Baja California Sur. Mexico. Ciencias Marinas 19{l):15-28. Garcia-Dominguez, F.. S. A. Garcia-Gasca & J. L. Castro-Ortiz. 1994. Spawning cycle of the red clam Megapitaria aurantiaca (Sowerby, 1831) (Veneridae) at Isla Espiritu Santo. Baja California Sur, Mexico. J. Shellfish Res. 13(2);417-423. Grant, A. & P. A. Tyler. 1983a. The analysis of data in studies of inver- tebrate reproduction. I. Introduction and statistical analysis of gonad indices and maturity indices. Int. J. Invert. Reprod. 6:259-269. Grant. A. & P. A. Tyler. 1983b. The analysis of data in studies of inver- tebrate reproduction. If. The analysis of oocyte size/frequency data, and comparison of different types of data. Int J. Invert. Reprod. 6:271-283. Jabbour, R. 1988. Etude de la variabilite genetique d'une espece perliere de Basse Califomie Sud, Mexique: Pinctada mazatlanica (Henley 1855). Memoire Diplome d'Etudes Approfondies. Universite Mom- pellier II, France. 49 pp. Jaramillo, R., J. Winter, J. Valencia & A. Rivera. 1993. Gametogenic cycle of the chiloe Scallop (Chlamys amandi). J. Shellfish Res. 12(1): 59-64. Keen, M. 1971. Sea Shells of Tropical West America. 2nd ed. Stanford University Press, Stanford. California. 1.064 pp. Luna, L. G. (ed.). 1968. Manual of histologic staining methods of the Armed Forces Institute of Pathology. 3rd ed. McGraw-Hill, New York. 258 pp. MacDonald, B. A. & R. J. Thompson. 1985. Influence of temperature and food availability on the ecological energetics of the giant scallop Placopecten magellanicus . II. Reproductive output and total produc- tion. Mar. Ecol. Prog. Ser. 25:295-303. Malachowski, M. 1988. The reproductive cycle of the rock scallop Hin- nites giganteus (Grey) in Humboldt Bay, California. J. Shellfish Res. 7(3):241-348. Mason-Suastegui, J. M. 1987. Evaluacion de cinco dietas microalgales en el crecimiento larval de Modiolus capax (Conrad, 1837) y Pinctada mazatlanica (Hanley, 1845). (Molusca Bivalvia). Tesis de Maestria. CICIMAR-IPN, La Paz, B.C.S. Mexico. 70 pp. Monteforte, M. 1990. Ostras perieras y Perlicultura; Situacion actual en los principales paises productores y perspectivas para Mexico. Serie Cienttfica U.A.B.C.S. (Numero Especial AMAC):13-18. Monteforte, M. & M. Cariiio. 1992. Exploration and Evaluation of natural stocks of Pearl Oysters Pinctada mazatlanica and Pleria sterna (Bi- valvia: Pteriidae): La Paz Bay, South Baja California, Mexico. Ambio 2l(4):3l4-320. Spawning Cycle of P. mazatlanica 303 Momeforte. M. & A. Garcia-Gasca. 1994, Spat collection studies on pearl oysters Pinchuki mazallaniai and Pterin sicrna (Bivalvia. Pteriidac) in Bahia de La Paz. South Baja California, Mexico, l-l\dnihii>lof>ia 291: 21-34. Monteforte. M. & H. Wright, 1944. Ecology of pearl oyster spat collection in Bahia de La Paz, South Baja California. Mexico; temporal and vertical distribution, substrate selection, associated species (Abstract: Pearls '94). J. Shellfish Res. 13(l):342-343, Rose, R. A., R. E. Dybdalh & S. Harders. 1990. Reproductive cycle of the Western Australian Silvcrlip Pearl Oyster Pinchula maxhnu (Jame- son) (Mollusca: Ptenidae). J. Shellfish Res. 9(21:261-272. Sastry, A. N, 1979, Pelecypoda (excluding ostreidae), pp. 131-192. In: A. C. Giese and J, S. Pearse (eds.). Reproduction of Marine Inverte- brates. Academic Press, New York. Saucedo, P. 1995. Crecimiento, relaciones alometricas y reproduccion de las ostras perleras Pinctada mazatlanica y Pteria sterna (Bivalvia: Pteriidae) bajo condiciones de repoblamiento en El Merito, Bahia de La Paz. Baja California Sur, Mexico. Tesis de Maestria, CICIMAR- IPN, La Paz, B.C.S. Mexico. 101 pp. Saucedo. P. & M. Monteforte. 1994. Breeding cycle of pearl oysters Pinctada mazatlanica and Ptena sterna in Bahia de La Paz. South Baja California, Mexico (Abstract: Pearls '94). J Shellfish Res. 13( 1 ):348- 349. Seed, R. 1976. Ecology, pp. 13-65. In: B, L, Bayne (ed.). Manne Mus- sels: Their Ecology and Physiology. Cambridge University Press, Cambridge. Sevilla. M. L. 1969. Contribucion al conocimiento de la madreperia Pinctada mazatlanica (Hanley, 1945). Rev. Soc. Mex. Hist. Nat. 30: 223-262. Signoret. M, & H. Santoyo. 1980. Aspectos ecologicos del plancton de la Bahia de La Paz, Baja California Sur. An. Centra Cienc. del Mar y Linuwl.. Univ. Nal. Aiitrin. Mexico. 7(2):217-248. Snedecor, G. W. 1950. Statistical Methods. Iowa State College Press, Ames, Iowa. 485 pp. Tranter, D. J. 1958a. Reproduction in Australian pearl oysters (Lamelli- branchia). I. Pinctada albina (Lamarck) Primary gonad development. Ausl. J. Mar. Freshwater Res. 9:135-143. Tranter, D, J, 1958b. Reproduction in Australian pearl oysters (Lamelli- branchia). II. Pinctada albina (Lamarck). Gametogenesis. Aust. J Mar. Freshwater Res. 9:144-158, Tranter, D. J. 1958c. Reproduction in Australian pearl oysters (Lamelli- branchial. III. Pinctada albina (Lamarck). Breeding season and sex- uality, Aust. J. Mar. Freshwater Res. 9:191-216. Tranter, D. J. 1958d, Reproduction in Australian pearl oysters (Lamelli- branchia), IV, Pinctada margaritifera (Linnaeus), Aust. J. Mar. Freshwater Res. 9:191-216. Tranter. D J 1959. Reproduction in Australian pearl oysters (Lamelli- branchia). V, Pinctada fi4cata (Gould). Aust. J. Mar. Freshwater Res. 10:45-56. Villalejo-Fuerte, M, & R. I. Ochoa-Baez. 1993. The reproductive cycle of the scallop Arfiopecten circularis (Sowerby, 1835) in relation to tem- perature and photoperiod. In Bahia Concepciiin, B.C.S., Mexico. Ciencias Marinas 19(2): 181-202. Villalejo-Fuerte. M., F. Garcia-Domi'nguez & R I Ochoa-Baez. 1995. Reproductive cycle of Glycymeris gigantea (Reeve, 1843) (Bivalvia: Glycymerididae) in Bahia Concepcion, Baja California Sur, Mexico. The Veliaer m2):\26-\n. Joiinml oj Shfllfish Reseunh. Vol, 13. No. 2. 305-.M1. 1996. OVERVIEW AND BIBLIOGRAPHY OF RESEARCH ON THE CHILEAN OYSTER TIOSTREA CHILENSIS (PHILIPPI, 1845) FROM NEW ZEALAND WATERS A. G. JEFFS' - AND R. G. CREESE^ ^Cawthron Institute Private Bag 2 Nelson, New Zealand 'Leigh Marine Laboratory University oj Auckland Private Bag 92019 Auckland. New Zealand ABSTRACT An overview of the biology and research on the Chilean oyster iTioslreii chilfiisis Philippi 1845) from New Zealand waters is provided along with a comprehensive bibliography This complements the bibliography prepared by Toro ( 1995) for the same species in South American waters. KEY WORDS: Chilean oyster. Tioslrea chilensis. bibliography. New Zealand flat oyster INTRODUCTION The Chilean oyster Tioslrea chilensis is one of four species of oysters commonly found in New Zealand waters (Jeffs 1995a). Extensive beds of this oyster fonned the basis of one of the first substantial commercial fisheries in New Zealand over 130 years ago (M.A.F.) 1975b). In more recent times, there has been intense interest in the commercial development of this oyster for enhance- ment and aquaculture. This species has been the subject of nu- merous research projects, with the results mostly being commu- nicated only in local publications. The purpose of this article is to draw together a bibliography of these publications and provide a synthesis of the current state of our knowledge for this species from New Zealand waters. TAXONOMY T. chilensis is known by a variety of common names in New Zealand, including tiopara. mud oyster, flat oyster, deep-sea oys- ter, dredge oyster, Foveaux Strait oyster. Bluff oyster, Stewart Island oyster, and southern rock oyster (Jeffs 1995b). The variety of common names in use is a reflection of the range of habitats and locations in which this species can be found. This species has also been known by a variety of taxonomic names since it was first described from New Zealand by Button (1873). A variety of geographic forms of the species were subse- quently named as separate species: for example, Ostrea hefforcli (Finlay 1928) for small and squat-shaped oysters found attached to rocks on the shores of Otago Harbour, and Ostrea charlottae (Fin- lay 1928) for individuals with broadly frilled shells often found in deeper water. Chanley and Dinamani ( 1980) proposed a new genus, Tiostrea, containing two species that had previously been referred to as Ostrea liilaria from New Zealand and Ostrea chilensis. the Chil- ean oyster or "ostra,"" from the Pacific Coast of South America. The new genus was proposed on the basis of the highly distinctive larval shell structure shared by the two species. Subsequently, Buroker et al. ( 1983) synonymised the two species as T. chilensis on the basis of similarities observed in their ecology, life history, and biochemistry. These workers also confirmed that O. charlot- tae and O. heffordi were ecomorphs of this same widespread spe- cies. This nomenclature now has widespread acceptance in New Zealand (Beu and Maxwell 1990). Therefore, we have continued to use T. chilensis (Philippi 1845) when referring to the native species of New Zealand. However, debate on the taxonomic status of this species of oyster has continued with no clear consensus emerging (see Harry 1985, Toro 1995). ECOLOGY T. chdensis reaches its greatest natural abundances in the colder waters in the southern parts of New Zealand, where adult oysters can reach densities of over 150 m~" in unexploited pop- ulations (Cranfield 1968a). Relatively extensive subtidal oyster beds have formed the basis of long-established commercial fish- eries in Foveaux Strait and in Tasman and Golden Bays (Cranfield 1975a, Cranfield 1975b, M.A.F. 1975b, Wame 1989). The pres- ence of these well-known southern fisheries has created a common misconception of a distinctly southern distribution for the species. However. Tiostrea has been found throughout much of New Zealand (Figs. 1 and 2) from intertidal sites to depths as great as 150 m (Beu and Maxwell 1990) and 549 m (Record A0910— Collection of the New Zealand Oceanographic Institute). The spe- cies has been observed to occupy a variety of habitats: attached to rocks and wharf piles in the intertidal (Morton and Millar 1973, Westerskov 1980); deeper water (60-100 m) and muddy substra- tum, such as off Otago Heads and Chatham Islands (Powell 1979); coarse sandy-pebble gravel bottom in waters of intermediate depths (20-50 m) (Cullen 1962). Tioslrea appears to be able to tolerate a wide range of water temperatures and salinities. Salinities in Foveaux Strait are typi- cally oceanic, ranging from 31 to 35 ppm, whereas oysters found in Stewart Island inlets can be subject to long periods of low- salinity water in the order of 3-5 ppm (Westerskov 1980, Buroker et al. 1983). Water temperatures in Foveaux Strait can range from 9 to 10°C in winter and 15 to 17°C in summer (Cranfield 1968b), whereas populations in northern areas, such as the Manukau Har- bour, are subjected to winter lows of 11°C and summer water temperatures as high as 27°C (Jeffs, unpublished data). In addition to the natural populations in New Zealand, a small wild population exists in Wales, Great Britain, after being introduced there from collections of oysters from Foveaux Strait in 1963 and again in 1966 (Walne 1974). The oyster was introduced for scientific ex- 305 306 Jeffs and Creese -r+ \ Hauraki Gull S25 S30 S35 S40 Wellington Harbour asman Bay A ^ _ Chatham Islands ^"^A - Otago Harbour Stewart Island A^^Foveaux Strait S45 in in ui o CO Ul S50 $ .. + Figures 1 and 2. Maps of New Zealand showing the location of offshore and coastal T. chilensis records, compiled from a number of collections. Key locations mentioned in this article are also identifled. Symbols: (•! Collection of the New Zealand Oceanographic Institute. (^) Collection of the National Museum of New Zealand. (□) Collection of the Auckland Institute and Museum. (■) Collections of members of the Conchology Section, A.l. & M. (A) Personal observations (A.G.J.). periments aimed at investigating the potential of alternative spe- cies for replacing exhausted beds of the native European flat oys- ter, Ostrea edulis (Linnaeus) (Taylor 1987. Utting 1987, Utting and Spencer 1992. Richardson et al. 1993). T. chilensis is also found as an almost ubiquitous fossil throughout New Zealand, usually deposited in Pliocene and Pleistocene rocks formed in past high-energy, shallow-water environments (Beu and Maxwell 1990). LIFE HISTORY CHARACTERISTICS The unusual reproductive behaviour of this oyster has attracted attention from some researchers (e.g.. Roughley 1929. Millar and HoUis 1963). T. chilensis is a protandrous hermaphrodite that breeds in the spring and summer once water temperatures rise above the lowest winter temperatures: 9-10°C in Foveaux Strait and Otago Harbour (Stead 1971a. Cranfield 1968b. Westerskov 1980. Buroker et al. 1983). above 14°C in Tasman Bay (Tun- bridge 1962). above 18°C in Wellington Harbour (Hollis 1963). and above 13°C in the Hauraki Gulf and Manukau Harbour (Jeffs, unpublished data). In Foveaux Strait, only a small proportion of the adult popu- lation spawns as females each summer (as few as 10-12%). whereas as many as 70-90% will develop male gonads (Cranfield 1975a). Females produce between 7.000 and 120,000 eggs that. once mature, are 300-350 |i.m in diameter (Cranfield 1975a & d). The eggs are thought to be fertilised in the inhalant chamber of the adult oyster and are then retained in the gills, where they continue to develop through to late-stage pediveligers (Hollis 1962, Hollis 1963, Cranfield 1968a, Stead 1971a). Estimates of the larval in- cubation period vary from 15 to 38 days and are thought to be related to water temperature (Hollis 1963, Stead 1971a, Wester- skov 1980). After the release from the parent oyster, the late-stage veliger larvae are ready to settle and the prodissoconchs range in length from 416 to 514 ^x.m. height from 318 to 400 jjim. and from 195 to 250 |jLm in depth (Cranfield 1979d. Chanley and Dinamani 1980). The highly distinctive larval shells of Tiostrea lack the posterior dorsal sulcus, umbones. and all hinge structures, which are commonly used as diagnostic features within the family Os- treidae (Chanley and Dinamani 1980. Beu and Maxwell 1990). The free-swimming larval life appears to be of only a few minutes duration (Cranfield 1968b. Stead 1971a). although some larvae may be released earlier and/or spend much longer in the plankton (Cranfield and Michael 1989). The growth rate of T. chilensis is extremely variable between individuals, between lo- calities, and between years. In Foveaux Strait, oysters may grow from freshly settled spat to a height of 5-20 mm in their first summer. 15-40 mm in the second. 25-60 mm in the third, and Overview of the Chilean Oyster in New Zealand 307 35-80 mm in the fourth (Stead 1971a). The oysters usually be- come sexually mature in their second or third year (Mollis 1963. Cranfield 1975a). FISHERIES AND AQUACULTURE PRODUCTION In prehistoric times. T. chilensis was harvested in small quan- tities by Maori in many parts of the country (Park 1969, Sullivan 1973. Ritchie 1980. Fox and Cassels 1983). Today, the customary Maori and recreational harvest of T. chilensis is minimal (M. F. Bull. pers. comm). Two commercial fisheries are based on beds of oysters found in Foveaux Strait and in Golden and Tasman Bays. Both fisheries rely on mechanical dredging to harvest oysters from the sea floor. Currently, most of these oysters are sold into the domestic market, which perceives the oysters as delicacy. Some small quantities of oysters, however, are currently exported from New Zealand to French Polynesia. Australia, and parts of Asia (N.Z. Fishing In- dustry Board export database). The management of these two commercial fisheries has been the main impetus for research on T. chilensis in New Zealand. Starting in 1906. a large number of surveys, fishery assessments, and studies of fishery biology were undertaken by government agencies (e.g.. Hunter 1906. Tunbridge 1962. Sorensen 1968. Stead 1971a & b. Street and Crowther 1973. M.A.F. 1974. Cran- field et al. 1993). The Tasman and Golden Bays fishery was historically the smaller of the two fisheries and is not so well researched. Annual landings from this fishery currently remain stable at about 500 tonnes, although landings have increased to nearly 700 tonnes in the past season (M.A.F. unpublished data). Annual landings from the Foveaux Strait fishery peaked during the late 1960s at 164,000 sacks; a sack may contain up to 70 dozen oysters and may weigh up to 79 kg (Cranfield I979d. Michael and Cranfield 1983, Cranfield et al. 1991). In 1986, an endemic spe- cies of Bonamia (a haplosporidian parasite) was first identified as the cause of major mortality of oysters in Foveaux Strait (Doonan and Cranfield 1992, Doonan et al. 1994). From that time, the disease spread throughout Foveaux Strait, leading to closures in the fishery in 1992. Surveys during the late 1980s and early 1990s indicated that the oyster population had fallen below 10% of the estimated virgin stock and was in danger of collapse through re- cruitment failure (Cresswell 1993. Cranfield et al. 1993). The devastating effect of Bonamia on the Foveaux Strait oyster beds led to an intensive series of disease surveys and related studies (e.g.. Hine 1986b. Hickman and Jones 1986. Dinamani et al. 1986. Dinamanietal. 1987. Hinc 1991a. Hine 1992d. Cranfield et al. 1991. Cranfield et al. 1993). Bonamia is also present in oysters from Tasman Bay. although it does not appear to have had any major effect on the commercial fishery (Hine 1992c). Other, less important parasites found in this oyster have also been the subject of further investigations (e.g.. Howell 1965. Howell 1966. Howell 1967. Jones 1981). Various aspects of the food quality ofTiostrea. including food value, stor- age, handling, and the unusual capacity of these oysters to accu- mulate some heavy metals have also been investigated in several studies (Malcolm 1927. Malcolm 1929. Brooks and Rumsby 1965. Brooks and Rumsby 1967. Thomas 1969. Nielsen 1975. Fenaughty et al. 1988). The collapse of the Foveaux Strait fishery and the increasing local price and demand for this oyster have led to greater efforts to culture this species (Smith et al. 1992). Difficulties in obtaining sufficient oyster spat and deaths of livestock due to Bonamia have hampered efforts to date. Consequently, the aquaculture and en- hancement of this species have not yet become well established in New Zealand (Smith et al. 1992. Hayden 1988. N.Z. Trade De- velopment Board 1989). More recent research efforts in New Zealand have focussed on developing Tioslrea for aquaculture and enhancement (e.g.. Smith et al. 1992, Hickman 1992a, Hickman 1992b, Hickman et al. 1988, Jeffs 1995b, Street 1995). In Chile, however, the species has been farmed since 1975 and a number of government-operated experimental stations produce seed oysters for enhancing natural oyster beds (Osorio 1979, Winter et al. 1984, Chanley and Chanley 1991, Aiken 1993). Aquaculture pro- duction in Chile has also been hampered by difficulties in spat supply (Lepez 1983, Valencia Camp 1990). DISCUSSION There is considerable potential for the commercial develop- ment of Tioslrea. given that it is widely recognised as a excellent eating oyster that can command premium prices over and above other cultivated oysters, such as the Pacific oyster Crassostrea gifias (Thunberg 1793). Workshops of scientists and commercial interests held in New Zealand have all concluded that the full potential for the aquaculture and fisheries development of this species would only be realised through further research (Smith et al. 1992. New Zealand Marine Farmers' Association, pers. comm.). Of critical concern was research that led to improved spat supply, more efficient culture practices, and an improved under- standing of the dynamics of the disease bonamiasis. Similar con- cerns have also been identified in Chile (Lepez 1983, Valencia Camp 1990, Toro et al. 1995a, Toro et al. 1995b). Despite numerous studies, the reproductive biology in this spe- cies, particularly the factors influencing larval production, remains poorly understood. Also, low levels of post-settlement survival are reducing the potential effectiveness of fishery enhancement and hatchery operations in New Zealand (Drunimond 1993a. New Zealand Oyster Company & M. F. Bull. pers. comm.). Further research in these areas would allow the development of hatchery techniques that would increase the spat production for aquaculture and fishery enhancement. In turn, this would create more oppor- tunities for selective breeding programmes to improve important attributes of broodstock, such as faster growth and disease resis- tance. Improved production efficiency will also come from a greater understanding of how the performance of oysters is af- fected by culture practices and environmental factors. The disease bonamiasis poses a continuing threat to Tioslrea aquaculture and fisheries, although active research into the disease has expanded in recent years. The results of these studies are beginning to clarify the nature of the relationship between the oyster and the Bonamia parasite, including the mechanisms that may be involved in triggering outbreaks of the disease (Hine 1996). From a general ecological perspective. Tioslrea is unusual in that it maintains a widespread distribution that includes an enor- mous range of habitats. To date, most research has been concen- trated on a few populations of commercial significance. Closer examination of other populations is likely to provide interesting and valuable results. For example, geographically isolated popu- lations of oysters may exhibit characteristics well-suited to aqua- 308 Jeffs and Creese ulture, such as faster growth, disease resistance, or higher fertil- ity, as has been seen in other species of molluscs (Gjedrem 1983. Newkirk 1983. Castagna and Manzi 1989. McShane and Naylor 1995). Such populations are also likely to show marked genetic differences, given that gene flow between distant populations may be limited when the larval life is often extremely short (Buroker et al. 1993). A number of scientists from around the globe are currently working on some of the research priorities outlined here. Recent cooperation between scientists from New Zealand and the United Kingdom has proved extremely fruitful (Utting and Hickman 1995). Further international cooperation can be expected to lead to greater gains in our knowledge of Tiostrea and should be actively encouraged. ACKNOWLEDGMENTS This work was supported by contract 402 with the New Zealand Foundation for Science. Research & Technology. Librar- ians at Auckland University's Biosciences and Leigh Libraries and the Ministry of Agriculture and Fisheries Library at Greta Point greatly assisted in locating publications for this bibliography. Mike Page from the New Zealand Oceanographic Institute. Bruce Marshall from the National Museum of New Zealand, and Bruce Hayward from the Auckland Institute and Museum kindly ar- ranged and assisted with access to their collections. Members of the Conchology Section of the Auckland Institute and Museum were generous in sharing their collections and knowledge of col- lection locations with the authors. LITERATURE CITED Aiken. D. 1993. Bivalve culture in Chile, World Acimuuluire 2M,A):(>-\9. Allen. R. L. 1979, A yield model for the Foveaux Strait oyster (Ostrea hitaha) fishery, Rapp. P. -v. Re'un Cons. Int. E.xplor. Mer. 175:70- 79, Allen. R. L. & H. J. Cranfield, 1979, A dredge survey of the oyster population of Foveaux Strait, Rapp. P.-v Re'un Cons. Ini. E.xplor. Mer. 175:50-62, Anonymous. 1987, New Zealand oysters under threat, Parasiiol Todiix 3(2):36, Auckland Regional Council 1992 Interlidal seafood resources of the Manukau Harbour, Environ. Plunninx Div. Tech. Publ. 15:1-200, Beu. A, G, & P. A, Maxwell, 1990. Cenozoic Mollusca of New Zealand, N.Z. Geol. Sitrv. Paleonlol. Bull. 58:336-341, Bioresearchers Ltd, 1989. Shellfish resources of the Taniaki River, An unpublished report to the Auckland Regional Water Board, Auckland Regional Water Board. Auckland, 50 pp. Booth. J, D, 1979. Common bivalve larvae from New Zealand: Pterlacea. Anomiacea, Ostreacea, N.Z. J. Mar. Freslmal. Res. 13( I): 131-139, Booth, J, D, 1983, Studies on twelve common bivalve larvae, with notes on bivalve spawning seasons in New Zealand, N.Z. J Mar. Freshnvl. Res. 17:231-265, Bradford, J. M.. H. J, Cranfield & K, P, Michael. 1991. Phytoplankton biomass in relation to the surface hydrography of southern New Zealand and possible effects on the food chain. N.Z. J. Mar. Freslmal. Res. 25:133-144, Brooks, R. R. & M, G, Rumsby, 1965, The biogeochemistry of trace element uptake by some New Zealand bivalves, Liinnol. Oceanogr. 10:521-527. Brooks, R. R, cS: M, G. Rumsby, 1967, Studies on the uptake of cadmium by the oyster Ostrea sinuata (Lamarck). Aust. J. Mar. Freshwat. Res. 15:53-61, Bull, M. F. 1971. A preliminary study of the feasibility of oyster farming from rafts in Kenepuru Sound: growth rates, spat settlement and the spawning season, BSclHons) Thesis, Victoria University. Welling- ton. 58 pp. Bull. M. F. 1993, Dredge oyster enhancement in Golden and Tasman Bays. Aquaculture Update 6:4-5. Buroker. N. E. 1985, Evolutionary patterns in the Family Oslreidae: larv- iparity vs. oviparity, J . Exp. Mar. Biol. Ecol. 90:233-247, Buroker. N. E., P. Chanley, H. J. Cranfield & P. Dinamani. 1983. Sys- tematic status of two oyster populations of the genus Tiostrea from New Zealand and Chile. Mar. Biol. 77:191-200. Castagna. M. & J. J, Manzi. 1989, Clam culture in North America: hatch- ery production of nursery stock clams, pp. 111-126, In: J. J, Manzi and M. Castagna (eds.). Clam Mariculture in North America. Devel- opments in Aquaculture and Fisheries Science, vol. 19. Elsevier Sci- entific, Amsterdam, Acjuacullure 19:1 1 1 — 126, Chanley. M, H, & P, Chanley, 1991, Cultivation of the Chilean oyster, Tiostrea chilensis (Philippi. 1845). pp, 145-151, In: Menzel, W. (ed.). Estuarine and Marine Bivalve Mollusk Culture. CRC. Boston. Chanley, P, & P, Dinamani, 1978, Larval development studies. Catch 5(10):28-3I, Chanley, P, & P, Dinamani, 1980. Comparative descriptions of some oyster larvae from New Zealand and Chile, and a description of a new genus of oyster, Tiostrea. N.Z. J. Mar. Freshwat. Res. 14:103-120, Cranfield, H J 1968a, An unexploited population of oysters, Ostrea lii- lariu Hulton, from Foveaux Strait, Part I, Adult stocks and spatfall distribution, N.Z. J . Mar. Freshwat. Res. 2:3-22, Cranfield. H, J, 1968b, An unexploited population of oysters. Ostrea lu- laria Hutton, from Foveaux Strait, Part II. Larval settlement patterns and spat mortality. N.Z. J. Mar. Freshwat. Res. 2:183-203. Cranfield, H. J. 1970. Some effects of experimental procedure on settle- ment of Ostrea lutaria Hutton. N.Z. J. Mar. Freshwat. Res. 4:63-69. Cranfield, H, J. 1975a, Bluff oysters 1 , N.Z. Na/. Her/mgc 5:1965-1970. Cranfield. H J, 1975b, Bluff oysters 2. N.Z. Nat. Heritage 5:1993-1999, Cranfield, H J, 1977, How many oysters are in Foveaux Strait? N.Z. Commercial Fishing 16(2):6-7, Cranfield. H, J, 1979a, 1978 a vintage year for oysters. Caich March: 6-7, Cranfield. H.J, 1979b, Area A oyster bed study underway. Catch May:9, Cranfield, H, J. 1979c. Strait oyster catch rate declines. Catch December: 9-10. Cranfield. H, J, 1979d, The biology of the oyster. Ostrea lutaria. and the oyster fishery of Foveaux Strait, Rapp. P.-v. Re'un. Cons. Int. E.xplor. Mer. 175:44-49, Cranfield. H. J, 1980, Foveaux Strait oyster catch still declining. Catch December: 12-15, Cranfield, H, J, 1985, The 1984 Foveaux Strait oyster season. Catch December: 12: 14- 16, Cranfield, H, J, & R, L, Allen. 1977. Fertility and larval production of oysters in an unexploited population of oysters Ostrea lutaria Hutton. from Foveaux Strait. N.Z. J. Mar. Freshwat. Res. 11:239-253, Cranfield. H. J. & R. L, Allen, 1979, Mark recapture surveys of Foveaux Strait dredge oysters {Ostrea lutaria) population, Rapp. P.-v. Re'un. Cons. Int. E.xplor. Mer. 175:63-69. Cranfield, H. J. & K, P. Michael. 1989, Larvae of the incubatory oyster Tiostrea chilensis (Bivalvia: Ostreidae) in the plankton of central and southern New Zealand, N.Z. J. Mar. Freshwat. Res. 23:51-60, Cranfield, H, J,, I, J, Doonan & K, P, Michael, 1991. Assessment of the effects of mortality due to Bonamia on the oyster population of Foveaux Strait in 1990 and the outlook for management in 1991, New Zealand Fisheries Assessment Research Document 91/18, MAF Fish- eries. Wellington. 36 pp, Cranfield, H, J,, 1, D, Doonan & K. P, Michael. 1993, Foveaux Strait Overview of the Chilean Oyster in New Zealand 309 oyster (Tioslrea chilensis) assessment 1993. New Zealand Fisheries Assessment Research Document 93/21. MAF Fisheries. Ministry of Agnculture and Fisheries. Wellington. 14 pp. Cresswell. P. 1993. Developments m the Blulf oyster fishery. Aqiiaciilniir Update 6:5. Cullen. D. J. 1962. The influence of bottom sediments upon the distribu- tion of oysters in Foveaux Strait. New Zealand. N.Z. J. Geol. Geo- phys. 5:271-275. Curtin, L. 1992. Ongrowing. //;.■ Farming the dredge oyster: proceedings of a workshop. MAF Occasional Publ. 7:12-15. Dinamani. P. 1971. Identification of oyster species competing with rock oysters for settlement space. N.Z. Fisheries Res. Div. Informalum Leaflet. 1:1-9. Dinamani. P.. P. M. Mine & J. B Jones 1986. The parasite in the dredge oyster. Catch 13(ll):4-6. Dinamani, P., P. M. Hine & J. B. Jones. 1987. Occurrence and charac- teristics of the haemocyte parasite Bonamia sp. in the New Zealand dredge oyster Tioslrea httaria. Dis. Aquat. Org. 3:37^4. Dinamani. P.. R. W Hickman. P M Hine. J. B. Jones & H. J Cran- field. 1987a. Report on investigations into the disease outbreak in Foveaux Strait oysters. Tioslrea liitariu. 1986-87. Ministry of Agri- culture and Fisheries Discussion Paper. MAF Fisheries, Wellington 32 pp. Dinamani, P., R. W Hickman, P. M Hine, J. B. Jones & H. J. Cran- field. 1987b. Bonamia strikes deeper into Foveaux Strait oysters. Catch 14(3):3-5. Doonan, I. J. & H. J. Cranfield. 1992. Foveaux Strait oyster (Tioslrea chilensis) assessment. 1992. New Zealand Fisheries Assessment Re- search Document 92/11. MAF Fisheries. Ministry of Agnculture and Fishenes. Wellington. 29 pp. Doonan. I. J., H. J. Cranfield & K. P. Michael. 1994. Catastrophic re- duction of the oyster, Tioslrea chilensis (Bivalvia: Ostreidae), in Foveaux Strait, New Zealand, due to infestation by the protistan Bo- namia sp. N.Z. J. Mar. Freshwal. Res. 28:335-344. Doonan. I J., J. L. McKoy, M. F. Bull & D. H. Stead. 1985. Scallop surveys in the Nelson and Marlborough regions. 1960-82. MAF Oc- casional Puhl. 18:1-65. Doonan. I. J.. H. J. Cranfield, P. M. Hine. D. O. Fisher & B. Wesney. 1992. Bluff oyster assessment, pp. 27-28. MAF Fishenes. Manne Research Annual Report 1991-1992. MAF Fisheries. Wellington. Drummond. K. L. 1987a. 500 tonne catch restnction on Nelson oysters. Catch May; 12. Drummond. K. L. 1987b. Dredge oyster catch reduction. Catch October: 20. Drummond. K. L. 1993a. Report on Tasman Bay dredge oyster produc- tivity trials. Central Fisheries Region Internal Report 23:1-19. Drummond, K. L. 1993b. The Challenger dredge oyster (Tioslrea chilen- sis) fishery. Draft Fishery Assessment Research Document. Central Fishenes Region Internal Report 93:1-1 1. Fenaughty, C. M., D. M. Tracey & J. W. Lock. 1988. Heavy metal and organochlorine concentrations in New Zealand aquatic fish, crusta- ceans, and molluscs. New Zealand Fishenes Data Report. 34. MAF Fisheries. Wellington. 44 pp. Finlay, H. 1928. The recent Mollusca of the Chatham Islands. Trans. N.Z. Inst. 59(2):232-286. Fleming, C. A. 1952. A Foveaux Strait oyster-bed. N.Z. J. Sci. Tech. 34:73-85. Fox, A. & R. Cassels. 1983. Excavations at Aotea. Waikato 1972-75. Rec. Auckland Inst. Mus. 20:65-106. Frazer, A. 1994. Foveaux Strait dredge oyster fishery. Seafood N.Z. 2(3): 25. Fujio, Y., R. Yamanaka & P. J. Smith. 1983. Genetic vanation in marine molluscs. Bull. Jap. Soc. Sci. Fish. 49:1809-1817. Gjedrem, T. 1983. Genetic variation in quantitative traits and selective breeding in fish and shellfish, pp. 51-72. In: N. P. Wilkins and E. M. Gosling (eds.). Developments in Aquaculture and Fisheries Science, vol. 33. Elsevier Science, Amsterdam. Grange, K. R. 1979. Soft-bottom macrobenthic communities of Manukau Harbour, New Zealand. N.Z. J. Mar. Freslmat. Res. 13:315-329. Harry, H. W. 1985. Synopsis of the supraspecific classification of living oysters (Bivalvia: Gryphaeidae and Ostreidae). The Veliger 28:121- 158. Hayden, B. J. 1988. Evaluation of species with potential for aquaculture in New Zealand. Ministry of Agriculture and Fisheries Publication. MAF Fisheries. Wellington. 48 pp. Hickman. R. W. 1972. Rock lobsters feeding on oysters. N.Z. J. Mar. Freshwat. Res. 6:641-644. Hickman. R. W. 1979. Oysters slow to grow. Catch May:ll. Hickman. R. W. 1985. "A parental problem." Catch 12:28. Hickman, R. W. 1987. Growth, settlement, and mortality in experimental farming of dredge oysters in New Zealand waters. N .Z. Fish. Tech. Rep 1,1-18. Hickman. R. W. 1992a. Research aims to rejuvenate flat oyster stocks. Aquaculture Update 1:3. Hickman. R. W. 1992b. Dredge oyster culture and spat production. MAF Fishenes. Manne Research Annual Report 1991-1992. New Zealand Ministry of Agriculture and Fisheries. Wellington. 37 pp. Hickman. R. W. 1992c. Spat production. In: Farming the dredge oyster: proceedings of a workshop. MAF Occasional Publ. l:l-\ 1 . Hickman, R. W. 1993a. Dredge oyster research at Mahanga Bay. Aqua- culture Update 6:3. Hickman, R. W. 1993b. Dredge oyster spat supplies. Aquaculture Update 6:2. Hickman, R. W. |993c. Rope culture of oysters. /l(/H(jn//m;c f/piAHc 6:4. Hickman, R. W. 1994. N.Z dredge oysters studied in Wales. Aquaculture Update 10:7. Hickman, R. W. & P. Dinamani. 1986. Bonamia in Foveaux Strait. Catch 13(lll:l. Hickman, R. W. & J. B. Jones. 1986. Foveaux Strait Oyster Disease Survey. Catch 13(11): 1-3. Hickman, R. W , D R Schiel. P. M Hine & M F. Bull 1988 Devel- opment of farming and enhancement techniques for the dredge oyster, Tiostrea lutaria. Strategic Planning for Aquaculture Report. Ministry of Agriculture and Fishenes, Wellington. 20 pp. Hine, P. M. 1986a. Bonamia: an overseas perspective. Catch 13( 1 1 ):2-4. Hine. P. M. 1986b. Interim report on studies of disease caused by a Bonamia-\\V.t organism in oysters (Tiostrea lutaria). Report to the Oyster Advisory Committee. MAF Fisheries. Wellington. 1 p. Hine. P. M. 1986c. Second interim report on studies of disease caused by a Bonamia-\\kc organism in oysters (Tiostrea lutaria). Report to the Oyster Advisory Committee. MAF Fisheries. Wellington. 2 pp. Hine, P M. 1987a. Third interim report on studies of disease caused by a Bonamia organism in oysters (Tioslrea lutaria). Report to the Oyster Advisory Committee. MAF Fisheries. Wellington. 2 pp. Hine, P. M. 1987b. Fourth interim report on studies of disease caused by i Bonamia organism in oysters (Tiostrea lutaria). Report to the Oyster Advisory Committee. MAF Fishenes, Wellington. 3 pp. Hine, P. M. 1987c. Fifth interim report on studies of di.sease caused by a Bonamia organism in oysters (Tioslrea lutaria). Report to the Oyster Advisory Committee. MAF Fishenes, Wellington. 5 pp. Hine, P. M. I987d. Sixth interim report on studies of disease caused by a Bonamia organism in oysters (Tioslrea lutaria). Report to the Oyster Advisory Committee. MAF Fisheries, Wellington. 2 pp. Hine. P. M. 1988a. Seventh interim report on studies of disease caused by a Bonamia organism in oysters (Tiostrea lutaria). Report to the Oyster Advisory Committee. MAF Fisheries. Wellington. 5 pp. Hine. P. M. 1988b. Eighth intenni report on studies of disease caused by a Bonamia organism in oysters (Tioslrea lutaria). Report to the Oyster Advisory Committee. MAF Fisheries. Wellington. 6 pp. Hine. P. M. 1988c. State oi Bonamia infection. Catch June:ll. Hine, P. M. 1989. Ninth interim report on studies of disease caused by a Bonamia organism in oysters (Tiostrea lutaria). Report to the Oyster Advisory Committee. MAF Fisheries, Wellington. 3 pp. 310 Jeffs and Creese lline, P. M, 1990a. Tenth interim repoil on studies of disease caused by a Bonamia organism in oysters (Tioslrea hilana). Report to tlie Oyster Advisory Committee. MAP Fisheries, Wellington. 6 pp. Hine, P. M. 1990b. Report on a survey of Bonamia in Foveaux Strait oyster beds: June-July 1990. Report to the Oyster Advisory Commit- tee. MAF Fisheries, Wellington. 12 pp. Hine, P. M. 1990c. Parasites and diseases of commercially important mol- luscs in New Zealand. Adv. Trop. Aquacullure . Aquacorp. IFREMER Actes de Colloque. 9:199-206. Hine, P. M. 1991a. The annual pattern of infection by Bonamia sp. in New Zealand flat oysters. Tiosrrea chilensis. Aqnaciilture 9.V.241-25I . Hine, P. M. 1991b. Ultrastructural observations on the annual infection pattern of Bonamia sp. in flat oysters Tiostrea chilensis. Dis. Aquat. Org. 11:163-171. Hine, P. M. 1992a. Bonamia update. Aquacullure Update 4:1. Hine, P. M. 1992b. Bonamia update, Aquacullure Update 1:3. Hine. P. M. 1992c. Disease control. In: Fanning the dredge oyster: pro- ceedings of a workshop. MAF Occasional Publ. 7:12-15. Hine. P, M. 1992d. Ultrastructural and ultracytochemical observations on Bonamia sp. in oysters (Tiostrea chilensis). with a consideration of organelle function. PAMAQ IV: Fourth International Colloquium on Pathology in Marine Aquacullure. 107(2-31:175-183. Hine, P, M. 1996. The ecology of Bonamia and decline of bivalve mol- luscs, N.Z. J. Ecol. 20: in press, Hine, P. M. & J. B. Jones. 1994. Bonamia and other aquatic parasites of importance to New Zealand, N.Z. J. Zool. 21:49-56. Hine, P, M. & B. Wesney. 1992. Inter-relationships of cytoplasmic struc- tures in Bonamia sp. (Haplosporidia) infecting oysters Tiostrea chil- ensis: an interpretation. Dis. Aquat. Org. l4(l):59-68. Hine, P. M. & B. Wesney. 1995a. The functional cytology of Bonamia sp (Haplosporidia) infecting oysters Tiostrea chilensis: An ultracytochem- ical study. Dis. Aquat. Org. 20(3):207-217. Hine, P M. & B. Wesney. 1995b. Interaction of phagocytosed Bonamia sp. (Haplosporidia) with haemocytes of oysters Tiostrea chilensis. Dis. Aquat. Org. 20( 3 ):2 19-229. Hollis, P. J. 1962. Studies on the New Zealand mud-oyster Ostrea lutaria Hutton. 1873. M.Sc. Thesis. Victona University, Wellington. 167 pp. Hollis. P. J. 1963. Some studies on the New Zealand oysters. Zool. Piihl Victoria Univ. 31:1-28. Howell. M. J. 1963. Theoretical lifecycle of a trematode parasite infecting Foveaux Strait oysters. N.Z. Commercial Fishing 1:16-18. Howell. M. J. 1965. Studies on the life history of Bucephalus longicor- nutus (Manter. 1954) (Trematoda: Bucephalidae) and the biological control of its sporocysts in Ostrea lutaria Hutton. M.Sc. Thesis. Vic- toria University. Wellington. Howell. M. J. 1966. A contribution to the life history of Bucephalus longicornutus (Manter. 1954). Zool. Publ. Victoria Univ. 40:1-42. Howell. M. J. 1967. The trematode. Bucephalus longicornutus (Manter. 1954) in the New Zealand mud-oyster, Ostrea lutaria. Trans. Rox. Soc. N.Z. Zool. 8(22):221-237. Hunter, R. C. 1906. Report on oyster beds in Foveaux Strait. Rep. Fish. N.Z. 1905:16-17. Hutton. F. W. 1873. Class Lamellibranchiata, Ostrea. Catalogue of the Marine Mollusca of New Zealand, with Diagnosis of the Species. Government Print, Wellington pp. 84-85. Jeffs, A. G. 1995a. The aquacullure and biology of oysters. Seafood N.Z. 3(2):30-31. Jeffs, A, G, 1995b. The aquacullure and biology of Bluff oysters. Seafood N.Z. 3(3):37-39. Jones, J, B, 1981, A new microsporidium from the oyster Ostrea lutaria in New Zealand, J. Invertebr. Pathol 38(l):67-70. King, M. 1980. Nelson oyster beds watched. Catch December:ll. King. M. & A. Branson. 1977. 1977 oyster season. Catch December:l7 King. M. & H. J. Cranfield. 1976. Bluff oyster seasons compared. Catch November:4-5. King, N.. M. Miller & S. de-Mora. 1989. Tributyl tin levels for sea water, sediment, and selected marine species in coastal Northland and Auck- land. New Zealand, N.Z. J. Mar. Freshnal. Res. 23(2):287-294. Kingett-Mitchell & Associates Ltd. 1988. Assessment of the status of the edible shellfish resources of the Manukau Harbour. An unpublished report to the Auckland Regional Water Board's Manukau Harbour Action Plan. Auckland Regional Water Board. Auckland. 48 pp. Lepez. Ml. 1983. El cultivo de Ostrea chilensis en la zona central y sur de Chile. Mems. Asoc. Latinoam. Acuicull. 5(2):1 17-127. Luckens, P. A. 1976. Settlement and succession on rocky shores at Auck- land, North Island, New Zealand. N.Z. Oceanogr. Inst. Memoir 10:\- 64. Makarios, E. 1994. Bluff oyster fishing and the Karaka. Seafood N.Z. September: 101-104. Malcolm, J. 1927. Food values of New Zealand fish. Part 8. Stewart Island oysters. Trans. N.Z. Inst. 58:167-173. Malcolm. J. 1929. Food values of New Zealand fish. Part 10. Seasonal variations in Stewart Island oysters. Trans. N.Z. Inst. 59:668-679. McShane, P. E. & J. R. Naylor, 1995, Small-scale spatial variation in growth, size at maturity, and yield- and egg-per-recruit relations in the New Zealand abalone Haliolis iris. N Z. J. Mar. Freslnvat. Res. 29: 603-612. Mialhe, E., V. Boulo, R. Elston, B. Hill, M. Hine, J. Monies. P. van- Banning & H. Gnzel. 1988. Serological analysis of Bonamia in Ostrea edulis and Tiostrea chilensis using polyclonal and monoclonal antibod- ies. Aquat. Living Resoiir. l(l):67-69. Michael. K. P. & H. J. Cranfield. 1983. Farming the Bluff oyster beds. Catch September: 16-1 7. Michael, K. P., 1. J. Doonan & H. J. Cranfield. 1990. Results of the Foveaux Strait oyster population pilot survey cruise JCO90/I0. MAF Fisheries Greta Point — Internal Report. 158:1-31. Millar, R. H. & P. J. Hollis. 1963. Abbreviated pelagic life of Chilean and New Zealand oysters. Nature 197:512-513. Ministry of Agriculture &. Fishenes. 1987. Foveaux Strait oysters surveyed again. Catch 14(1 ):3. Ministry of Agriculture & Fisheries. 1974. Dredge oyster shells returned. CarWi July:5. Ministry of Agriculture & Fishenes. 1975a. 1974 shellfish production — dredge oysters. Catch August:20. Ministry of Agriculture & Fisheries. 1975b. History of the dredge oyster fishery. Catch September:3-4. Ministry of Agnculture & Fishenes. 1975c. Summary of the 1975 Bluff oyster season. Catch September:4— 5. Ministry of Agriculture & Fisheries. 1976. 1976 Foveaux Strait oyster quota set. Catch February:3. Ministry of Agnculture & Fishenes. 1979. Farmed Bluff Oysters. Catch November:34. Ministry of Agnculture & Fisheries. 1987. Bluff oyster management re- gime to continue. Calch July:9-10. Ministry of Agnculture & Fisheries. 1992. Dredge oyster workshop. Aquacullure Update 3:1. Ministry of Agriculture & Fisheries. 1994. Aquacullure and Fisheries En- hancement; dredge oysters and shellfish diseases. MAF Fisheries Ma- rine Research, Annual Report for 1992-93. New Zealand Ministry of Agriculture and Fisheries, Wellington, pp. 39-40. Morton, J. E. M. & M. C. Millar. 1973. The New Zealand Sea Shore. 2nd ed. Collins, Auckland. 653 pp. Newkirk, G. F. 1983. Applied breeding of commercially important mol- luscs; a summary of discussion, pp. 415—422. In: N. P. Wilkins and E. M, Gosling (eds.). Developments in Aquacullure and Fishenes Sci- ence, vol, 33. Elsevier Science. Amsterdam, Aquacullure 33:415—122. New Zealand Trade Development Board. 1989, Directions in Foreign Ex- change Earnings; the New Zealand Aquacullure Industry. Wellington. 55 pp. Nielsen. S. A. 1973. Effect of acetazolamide on larval settlement of Os- trea lutaria. The Veliger 16:66-67. Nielsen. S. A. 1975. Cadmium in New Zealand dredge oysters: geo- graphic distribution. Intern. J. Environ. Anal. Chem. 4:1-7. Mordberg, M., I. Nuottaniemi, M. G. Cherian, G. F, Nordberg, T. Kjell- Overview of the Chilean Oyster in New Zealand 311 stroem & J S, Garvey. I98h. CharactcriMituin studies on the cad- mium-binding proteins from two species of New Zealand oysters l.n- viniii. Hfiillli Perspea. 65:57-62. Osorio. R. C. 1979. Moluscos marinos de importancia economica en Chile. Biot. Pesq. Chile 11:3-47. I'ark. G, S. 1969. Tiwai Point — a preliminary report. N.Z. ArchueoL As- soc. Newsletl. 12(3): 143- 146. Pickering. T. 1992. Gazette notice to control the spread of Bciuimid. Aquaculture Updaw 4:1. Powell. A. W. B. 1979. New Zealand Molliisca: Marine. Land and Fresh- water Shells. Collins. Auckland. 500 pp. Richardson. C. A.. R. Seed. E. M. H. Al-Roumaihi & L. M. McDonald. 1993. Distribution, shell growth and predation of the New Zealand oyster. Tiostrea ( =OslrcLi) lularia Button, in the Menai Strait. North Wales. J. Shellfish Res. 12:207-214. Ritchie. N. 1980. Archaeology and Prehistory of the Upper Wailaki Re- gion. Otago Museum Annual Report 1965-1966. 18 pp Roughlev, T, C. 1929. Monoecious oysters. Nature 124:793. Shamia. R. P. 1983. Ligands binding cadmium, zinc and copper in a species of New Zealand oyster {Oslrea lularia). Bull. Emiron. Con- lam. To.xicol. 30(41:428-434. Smith. P. J.. G. G. Baird & M. F. Beardsell (eds.). 1992. Farming the dredge oyster: proceedings of a workshop July 1992. N.Z. Fisheries Occasional PiM. 7:1^3. Sorensen. J. H. 1968. Dredge oyster surveys Foveaux Strait pre-1960. N Z. Fish. Tech. Rep. 28:64 pp 4- appendices and illustrations. Stead. D. H. 1962. Preliminary report on the Foveaux Strait oyster inves- tigation 1960-62. N.Z. Mar. Depl. Spec. Fisheries Rep. 16:1-28. Stead. D. H. 1971a. Observations on the biology and ecology of the Foveaux Strait dredge oyster {Oslrea lularia Button). iV.Z. Fish. Tech. Rep 68:1-49 Stead. D. H. 1971b. Survey of Foveaux Strait oyster beds I96(.)-1964. .V Z Fish. Tech. Rep. 59:64 pp -I- appendices. Stead. D. B. 1973. Shellfish survey Nelson sewer outfall area April-June 1972. N.Z. Fish Tech. Rep. 123:1-15. Street. R. J. 1995. Oyster enhancement trials in Foveaux Strait. Seafood N.Z. 3:32-34. Street. R. J. & G. S. Crowther. 1973. Foveaux Strait oyster surveys 1965-1971. N.Z. Fish. Tech. Rep. 118:1-14. Street. R. J. & S O'Ballorane. 1985. Foveaux Strait oyster beds— effects of bottom trawling studied. Calch 12(3):16-I7. Street, R. J.. Crowther. G. S. & J. S. Kirkman. 1973. Oyster shell return experiment — Foveaux Strait. N.Z. Fish. Tech. Rep. 107:1-16. Sullivan. A. 1973. Site survey of Lower Pukaki Creek, vol. 27. Working Papers in .Anthropology. Archaeology. Linguistics and Maori Studies; Working Paper in Archaeology. Department of Anthropology. Univer- sity of Auckland. Auckland. 94 pp -i- appendices. Taylor. B. 1987. N.Z. oyster could help revive U.K. oyster farming. Catch August:9. Thomas. J. 1969a. A study of factors affecting the frozen storage life of New Zealand deep sea oyster {Oslrea lularia). N.Z. J. Sci. 12:777- 783. Thomas. J. 1969b. Correlation of pH and quality of fresh New Zealand oysters {O.'.lrea lularia). N.Z. J. Sci. 12:784-788. Toro. J. E. 1995. A bibliography of literature on the Chilean oyster Oslrea chilensis (Philippi. 1845). J Shellfish Res. 14:93-96. Toro. J. E.. M. A. Sanhueza. J. E. Winter. P. Aguila & A. M. Vergara. 1995a. Selection response and heritabilily estimates for growth in the Chilean oyster Oslrea chilensis (Philippi. 1845). J. Shellfish Res. 14(l):87-92. Toro. J. E.. M. A. Sanhueza. J. E. Winter. C. M. Senn. P. Aguila & ■A. M. Vergara. 1995b. Environmental effects on the growth of the Chilean oyster Oslrea chilensis in five mariculture locations in the Chiloe Island. Southern Chile. Aquaculture 136:153-164. Tunbridge. B. R. 1962. Occurrence and distnbution of the dredge oyster (Oslrea sinuala) in Tasman and Golden Bays. N.Z. Fish. Tech. Rep. 6:1-42. Utting. S. 1987- Fresh prospects for bivalve farming. Fish Farmer 10:17- 18. Utting. S. & R. W. Bickman. 1995. Cooperative dredge oyster studies. Aquaculture Update 11:4. Utting. S. D. & B. E. Spencer. 1992. Introductions of marine bivalve molluscs into the United Kingdom for commercial culture — case his- tories. Introductions and Transfers of Aquatic Species. Selected papers from a symposium held in Halifax. Nova Scotia. June 1990. Sinder- mann. C. Steinmetz. B. & W. Bershberger (eds.). 194:84-91. Valencia Camp. J. 1990. Eslado actual del cultivo de ostra chilena. Tios- trea chilensis. In: Proc. 6th Taller de Acuacultura — Actualidad. tec- nologia. mercado e interaccion social del cultivo del ostiiin. ostra y especies potenciales. Universidad Catolica del Norte y Asociacion de Productores de Ostras y Ostiones pp. 99-109 Voller. R. 1988. Foveaux Strait oyster season quota for 1988. Catch March: 18-19. Walne. P. R. 1970. Studies on the food value of nineteen genera of algae to juvenile bivalves of the genera. Oslrea. Crassosirea. Mercenaria and Mxtilus. Ministry of Agriculture. Fisheries and Food. London. Fishery Invest. Series II 26(5):l-62. Walne. P. R. 1974. Culture of Bivalve Molluscs. 50 years' expenence at Conwy. Fishing News Books Ltd. Famham. Surrey. England. 189 pp. Walsby. J. 1989. The cultured oyster. N.Z. Geographic 3:70-81. Wame. K. 1989. Bluff gold. N.Z. Geographic 3:62-69. Waugh. G. D. 1969. Dredge oyster farming. Fisheries Res. Div. Occa- sional Puhl. 3:1-;. Westerskov. K. 1980. Aspects of the biology of the dredge oyster Oslrea lularia Button. 1873. Ph.D. Thesis. University of Otago. Dunedin. 192 pp. Winter. J.. J. Toro. J. Navan-o. G. Valenzuela & O. Chapan-o. 1984. Recent developments, status, and prospects of molluscan aquaculture on the Pacific coast of South Amenca. Chapt. 14. vol. 39. In: D. Morse. K. K. Chew & R. Mann. (eds.). Recent Innovations in Culti- vation of Pacific Molluscs. Developments in Aquaculture and Fisheries Science. Elsevier Scientific. Amsterdam. 95 pp. Journal of Shellfish Research. Vol. 15. No. 2. 313-318. 1996. ENHANCEMENT OF SUBTIDAL EASTERN OYSTER, CRASSOSTREA VIRGIN IC A, RECRUITMENT USING MESH BAG ENCLOSURES FRANCIS X. O'BEIRN,' * RANDAL L. WALKER,' AND PETER B. HEFFERNAN- ^Universin- of Georgia Shellfish Aquaculture Laboratory 20 Ocean Science Circle Savannah. Georgia 'Marine Institute 80 Harcourt Street Dublin 2, Ireland ABSTRACT Eastern oysters, Crassoslrea virginica. in the southeastern United States are found predominantly in the intertidal zone. In this study, mesh bags (3 and 6 mm) were deployed over colleclmg frames, and the patterns of oyster settlement on these collectors were compared against unmeshed controls at three tidal heights (intertidal, low water, and subtidal) over three sampling regimes (biweekly, monthly, and seasonal) at two sites. Within the biweekly sampling regime, the meshed collectors and controls had similar patterns of settlement at the respective tidal heights. For monthly samplers, mesh treatments maintained higher settlement suhtidally whereas controls had highest settlement on the collectors at mean low-water level. Controls had highest recruitment intertidally for seasonal collectors, whereas mesh treatments had higher recruitment lower in the intertidal zone. Conclusions from this expenment were that the use of mesh-covered collectors enhanced subtidal oyster recruitment. Causes of observed increases in subtidal settlement in mesh collectors over unmeshed controls over time could be the result of a combination of factors; predator exclusion, larval entrainmenl. or reduced desiccation, which seemed to overcome the detrimental effects of increased fouling, resulting in reduced flow and possible hypoxic conditions within the mesh bags. Given the degree of recruitment and the sizes of the recruits attained w ithin the mesh bags, the use of these methods to attain juveniles for commercial purpo.ses would appear to be both feasible and viable, particularly for long periods (up to 6 mo) of deployment. KEY WORDS: Crassoslrea virginica. mesh excluder, oysters, predation, recruitment INTRODUCTION The range of eastern oyster, Crassostrea virginica (Gmelin), extends from the Gulf of St. Lawrence south along the eastern seaboard of the United States and throughout the Gulf of Mexico (Galstoff 1964). Within this range, the oyster is found predomi- nantly in the subtidal zone. However, in the southeastern United States, and specifically in South Carolina and Georgia, the ma- joiity of oysters are intertidal. The primary reasons given for the lack of subtidal oysters have been disease, mismanagement of the resource, competition from other epibionts. and predation (Harris 1980. Ofiara and Stevens 1987. Michener and Kenny 1991). Macropredators are numerous and include mammals, fish, crustaceans, and molluscs (Linton 1968, Walker 1981. Walker 1993). However, relatively little is known concerning the effect of these predators on newly settled oysters. Predation on young oysters by blue crabs, Callinectes sapidus. was identified as a contributing factor in the 1946 failure of oyster recruitment in South Carolina (Lunz 1947). Galstoff (1964) observed that many young oysters were adversely affected by crabs feeding on larger oysters on which larvae had settled and attached. Recruitment studies by O'Beim et al. (1995 and 1996) sug- gested that events shortly after oyster settlement and metamorpho- sis may contribute to the confinement of oysters intertidally in coastal Georgia. It was observed that oyster numbers recorded on 'Corresponding Author; Present Address; Dept. of Fisheries and Wildlife Sciences, Virginia Polytechnic Institute and State University, Blacks- burg. VA 24061-0321. collectors were significantly higher in the subtidal than intertidal environment over short periods of sampler deployment (i.e.. 2 wk and 1 mo), with the reverse observed for longer periods (up to 7 mo). Furthermore, it was suggested that with greater duration of deployment and consequently the submergence of subtidal collec- tors, the potential for predation or other mortality factors (such as competition or exposure to pathogenic organisms) on oysters was increased. We describe a field study to evaluate oyster settlement using protective mesh bags that in theory would allow oyster lar- vae to set on collectors while limiting the access of potential pred- ators. Also, we anticipated that the use of protective meshes would increase the potential to harvest adequate numbers of spat from the subtidal zone in the southeastern United States. STUDY SITES The work was carried out from April to November 1993 at two sites and April to September 1994 at one site (for logistical rea- sons), in Wassaw Sound, GA (Fig. 1). The two sites were; House Creek, a sheltered tidal creek near the mouth of the sound; and Skidaway River, a less sheltered site located on the Intracoastal Waterway. A more detailed description of the hydrographic char- acteristics of the sites can be found in O'Beim et al. ( 1995). The two sites have been used to monitor oyster recruitment since 1991 (O'Beirn 1995. O'Beim et al. 1995. O'Beim et al. 1996). METHODS The sampling apparatus and methods of deployment were sim- ilar to those already used in the established monitoring programs in coastal Georgia (O'Beim 1995, O'Beim et al. 1994. O'Beim et al. 1995. O'Beirn et al. 1996). Briefly, longitudinally grooved poly- 313 314 O'Beirn et al. Figure 1. Map of sampling area in Wassaw Sound. GA, indicating the two sampling sites used throughout the study: (1) House Creek and (2) Skidawav River. vinylchloride (PVC) tubing embedded with chips of calcium car- bonate was used for collecting spat. A 12-cm section of tubing on each collector provided a sampling area of approximately 100 cm". Three collectors were arranged vertically on a sampling unit; each collector was exposed to one of three tidal heights (subtidal. mean low water, and intertidal). Four replicate sampling units were attached to a portable frame, which in turn was attached to a fixed frame (Fig. 2). After return to the laboratory, each collector was rinsed to remove extraneous material and exammed with a binocular microscope at 10 x to enumerate the number of oysters on each collector. Collecting frames with four replicate sampling units were cov- ered with 3- and 6-mm mesh bags (see Fig. 2). The open end of each bag was sealed, folded, and inserted into a PVC pipe that had been slit longitudinally. Hereafter, the 3- and 6-mm mesh bag- covered frames shall be referred to as the 3- and 6-mm treatments, respectively. An exposed control frame was also placed at each site. Collectors were retrieved and replaced at the two sampling sites on a biweekly and monthly basis. Thus, every 2 wk and once monthly, three frames (i.e.. two mesh treatments and one control) were taken from each site and evaluated for spat. A separate set of Lowwater Collector Subtidal - Collector Figure 2. Schematic diagram of the sampling apparatus used in the study. For the 3- and the 6-mm treatments, mesh bags were slipped over the sample frame. seasonal collectors was also deployed at the two sites. These were left on site for the duration of the study. These seasonal collectors gave an estimate of the overall recruitment of oysters for the entire spawning season. The experiment was repeated in 1994 at the Skidaway River site only. STATISTICAL ANALYSIS The enclosure of each of the treatment frames within a mesh ■"bag" would result in some problems of analysis relating to pseu- doreplication (Hurlbert 1984). Therefore, data at each tidal height, within each sampling period, were pooled and each period was then used as a replicate with which to carry out the analysis. The substantial variation in oyster numbers retrieved on the collectors throughout the study necessitated the log transformation [In(x -I- 1)1 of these values within each of the data sets retrieved. To evaluate the patterns of settlement within each treatment, one-way analysis of variance (ANOVA) and the Tukey Studentized Range Test (when appropriate) were pert'ormed on these transformed data with tidal height as the main effect. Separate ANOVAs were per- formed on the data for each mesh treatment, within each sampling regime at each site, and in the case of the Skidaway River, each year. The primary goal was to evaluate the patterns of settlement within each treatment. It was hypothesized that the mesh treat- ments would retain proportionally greater numbers of spat, sub- tidally over time, than the unmeshed control. A standard signifi- cance level of 5% was chosen for all statistical tests (a = 0.05). RESULTS Overall, oyster recruitment in Wassaw Sound in 1993 was considerably lower than that in 1991 and 1992 (O'Beirn 1995. O'Beirn et al. 1996). At House Creek, recruitment on the bi- weekly control collectors was first observed in late May 1993. However, peak settlement did not occur until late September (x = 7.4 spat/0.01 m-; Fig. 3A). At the Skidaway River site in 1993. levels were extremely low until peak settlement in late September (X = 63.2 spat/0.01 m"; Fig. 3B). Settlement was high on the monthly control collectors at House Creek early in the 1993 sea- son, after which, settlement dropped off and showed no apprecia- ble increase over the rest of the season (Fig. 4A). Peak monthly settlement on the intertidal collectors occurred in May at the House Creek site (x = 225.6 spat/0.01 m"). Similar patterns were ob- served at the Skidaway River site; peak settlement occurred early in the 1993 season, followed by low settlement levels throughout the year. Peak settlement at the Skidaway site was on the low- water collectors in June (x = 154.4 spat/0.01 m'; Fig. 4B). Oyster settlement on the biweekly subtidal collectors at the Skidaway River site in 1994 commenced in mid-May and increased to a peak in mid-June (x = 5.75 spat/0.01 m"; Fig. 38). Monthly settlement values had peak settlement on the low-water collectors in July (X = 26.3 spat/0.01 m"; Fig. 4B). Results of the ANOVAs on the transformed biweekly, monthly, and seasonal data from House Creek and Skidaway River are presented in Table 1 . For the biweekly data at both sites in 1993 and at Skidaway in 1994 (Table 1), controls tended to retain greater number of oysters than both of the treatments. The lack of significance of many of the tests indicated no distinct patterns. The highest settlement was achieved on subtidal and low-water collectors for both treatments and controls at Skidaway in 1994 (Table 1). SuBTiDAL Oyster Recruitment 315 Biweekly A. House Creek Tidal Height B. Skidaway River 1994 ""8 iepT~< l''is"'300 CFU/mL). t NA. not aoDlicable. Juvenile Oyster Disease 321 = 400) and shell growth (n = 100-400). Statistical analysis of mortality patterns among cohorts was performed by the use of one-way ANOVA and a posteriori comparisons to ascertain the effects of exposure time on percent-transformed data. Critical F values are indicated when the differences are significant. Differ- ences in shell heights among cohorts at the end of the study were also analyzed by the use of one-way ANOVA (Sokal and Rohlf 1981. Glantz 1992). Sample Processing Oyster samples were cleaned and prepared according to meth- ods modified from the Recommended Procedures for the Exami- nation of Sea Water and Shellfish (APHA 1970). Oysters were homogenized as 10-g whole-oyster samples or as the meat and juice only of 12 oysters, depending on the size of the animals (Table 1). Homogenized oyster samples were then diluted deci- mally in sterile phosphate-buffered saline water (PBSW). Ten mil- liliters of each dilution of oyster homogenates was filtered through a sterile 0.45-n.m-pore-size cellulosic membrane filter in triplicate by aseptic techniques. Vibrio trapped on the membranes were preenriched by placing the membranes on absorbent pads soaked with sterile alkaline peptone water for 6 h at 30°C. Membranes were then aseptically transferred onto thiosulfate-citrate-bile salts (TCBS) agar plates and incubated for 18-24 h at 30T (Bryant et al. 1986, Venkateswaran et al. 1989). In the presumptive phase, the total number of colonies, most likely Vibrio spp. colonies, was enumerated on a colony counter to obtain the total number of colony-forming units (CPUs). In the confirming phase, selected colonies appearing in the highest (i.e., most abundant) dilutions were streaked onto new TCBS agar plates to isolate pure colonies from possible mixed colonies (Brock and Madigan 1991 ). Distinc- tive colony morphotypes were selected and transferred to marine agar (Difco® 2216E) for storage and subsequent testing. These isolates were also streaked onto tryptic soy agar plates for further identification by use of the Biolog Inc. (Haywood, CA), and API 20E (BioMerioux, France) diagnostic systems, as well as other biochemical and physiological tests conducted according to Ba- lows (1974), Cowen (1974), Difco (1991) and Holt et al (1994). Bacteria in sediment samples were dissociated from the parti- cles before analysis by the vigorous vortexing of 0.5-g subsamples in 10 mL of sterile sodium metaphosphate (25 mM; pH = 7.0) for 15 s and then the addition of 25 mL of sterile distilled water before being vortexed again for 15 s (modified from Saad and Cabelli, unpub. obs.). Suspensions were allowed to stand in a refrigerator, and then supematants were diluted as required in PBSW. Sediment sample dilutions were then subjected to the same preenrichment, presumptive, enumeration, and confirming protocols as described for oyster samples. Debris (mostly biodeposits) collected from the external surface of oyster shells in nursery growout trays were treated like sediment samples. Water samples were also filtered and treated like oyster samples for bacteriological analysis. Three batches of 12 individuals from cohort III oysters collected on October 3, 1993, were assayed separately to examine the parti- tioning of Vibrio within the oysters. Batch 1 consisted of whole oysters crushed, batch 2 consisted of pallial fluid only, and batch 3 consisted of meat, palliall fluid, and shells. From batch 3, meat and pallial fluid were assayed together, whereas shells were crushed and assayed separately. All samples were subjected to the same procedures as regular oyster samples. Similarities in Vibrio spp. concentrations among cohorts were analyzed by the use of regression analysis. Challenge Experiments Juvenile oysters used in challenge experiments were obtained from the East Hampton Hatchery, Long Island, NY. for the first experiment (avg. height, 29.3 mm; ±0.22 SE) and from Haskin Shellfish Research Laboratory, Rutgers University, NJ, for the second (avg. height, 35.4 mm; ±0.25 SE) and third (avg. height, 17.3 mm: ±0.20 SE) experiments. JOD has never been reported from these sites. Animals were about 1 (first and second challenge experiments) or less than 1 y old (third challenge experiment). Two weeks before the expenments, oysters were acclimated to the experimental temperature in flow-through ambient seawater in a ~30-L (20-26 ppt salinity) aquarium supplemented with cultured agae, Thalassiosira weissflogii or Isochrysis gatbanu. During challenge experiments, batches of oysters were maintained in in- dividual aquaria (~11-L capacity) in filtered seawater with con- stant aeration on a temperature-controlled seawater table at the State University of New York at Stony Brook, Flax Pond Facility. A total of nine Vibrio spp. isolates were used in the challenge experiments. Three selection criteria were used to increase the probability of selecting pathogenic bacteria from 200 isolates. The nine species of Vibrio were isolated from oyster meats to minimize the selection of nonpathogenic Vibrio spp. from the general aquatic surroundings. They were also chosen during periods pre- ceding the onset of high mortality by 1 to 3 wk in order to mini- mize the selection of strains that might be opportunistic secondary invaders of dead and dying oysters. A third criterion was the phenotypic resemblance to Vibrio PI (because of the close simi- larities between JOD and BRD). At least the first two criteria were met by all nine isolates. Small notches were cut into the ventral margin of the oyster shell edge with a Dremel tool. Random lots of —30 animals were then segregated for each treatment, and each treatment was dupli- cated. From each experimental animal, pallial fluid was removed and replaced with a 50-n,L inoculum of a Vibrio isolate in station- ary growth phase (—10* cells per injection), as described in Pail- lard and Maes ( 1990). Tank water was changed every 7 d. In the first and second challenges, three negative controls were run si- multaneously; (1) notched but no injection (blank), (2) filtered seawater injection (FSW), and (3) injection with a viable non- pathogenic bacteria, Escherichia cob (Difco Bactro Disks ATCC 25922) at the same cell density (E. coli). In the third experiment, negative controls were; (I) no notching and no injection (blank), (2) notched and no injection (notch), and (3) E. cob injection (E. coli). Each replicate treatment was placed in a separate aquarium containing 11 L of freshly filtered (0.2-|jLm-pore-size filter), aer- ated seawater. All aquaria were placed on a temperature-controlled seawater table. Oysters in the first challenge experiment received a single in- oculation of Vibrio spp., were maintained at 22°C under relatively uncrowded conditions (30-33 oysters per tank), and were batch fed twice daily (-66 x lO"" cells tank^' or 2.0 x 10*" cells oyster" ' of T. weissflogii, or - 19 x 10** cells tank " ' or ~6.0 x 10^ cells oyster" ' of/, galbana at each feeding) for 30 d. Biode- posits were removed daily with a pipette. Mortality was recorded daily, and dead oysters were removed. The second challenge ex- periment was run for 14 d under more stressful conditions. The animals were injected twice (days 0 and 7), fed only once a day, supplemented with 15 mL tank" ' of Vibrio suspension (10*" cells mL"' final tank concentration) every other day, and crowded inside I -mm mesh bags (9 x 10 cm enclosed area). Biodeposits 322 Lee et al. DATE Figure 1. Mean (±SE) cumulative mortalities of three experimental oyster cohorts and cohort II grown in 1- and 6-mm mesh growout trays (inset) of duplicate samples from duplicate trays at the Frank M. Flower and Sons (FMF) Co. oyster nursery. Oyster Bay. NY, during 1993. The dotted lines represent estimated total mortality. iL] Cohort I. (D) Cohort II, 1-mm mesh trays. (■) Cohort II, 6-mm mesh trays. (O) Cohort III. Error bars smaller than the symbols are not indicated. were not removed, and aquaria were maintained at a higher tem- perature of 24°C. At the end of the second experiment, oysters from the three treatments exhibiting highest mortaUty. as well as those from the three control batches, were processed and analyzed for bacterial abundances according to the protocols followed for field samples. The third challenge experiment was run for 21 d under a dif- ferent set of conditions. The animals received a single injection, were maintained at 25°C. fed twice a day. and were supplemented with Vibrio spp. suspension every other day. Oysters were crowded inside 1-mm mesh bags suspended in the water column, and the bottom of the aquaria contained sediment from Oyster Bay (175 g of dry weight sediment per tank, autoclaved 8-10 h at 12rC at 15 lb of pressure). Sediments were mixed with appro- priate Vibrio spp. isolates before oysters and filtered seawater were added into each tank. Biodeposits were removed, and tank water was changed every 7 d. In this experiment, the two isolates that induced the highest mortality in previous experiments were used separately and in combination. Two other new isolates were also used separately and in combination. All treatments from the third experiment, including controls, were processed for total bac- terial abundances and identification. To determine whether the bacteria that had been inoculated proliferated and thus satisfy Koch's postulates (Anderson and Sobieski 1980). the reisolated bacteria were identified and compared with the original inocula in the second and third experiments. One-way ANOVA and a priori multiple comparisons between controls and experimental batches, as well as a posteriori multiple comparisons among experimental batches and among all of the injected batches, were used to as- certain the effects of Vibrio injection. Critical F values are indi- cated when the differences were significant. RESULTS Oyster Mortality Patterns Mortalities were negligible in all three cohorts until mid-July. They were first recorded in cohort I on July 1 3. in cohort II on July 20, and in cohort III on August 10 (Fig. 1). Mortalities peaked at 21, 35, and 60% for cohorts I. II. and III, respectively (repre- sented by dotted lines in Fig. 1). 3-5 wk after mortality was first observed. JOD involvement in the mortalities was established by the presence of ridge-like conchiolin deposits on the inner shell (Fig. 6 in Bricelj et al. 1992). However, both conchiolin deposits and oyster mortalities were less severe than in previous years at this site. Only 0-407f of the population exhibited light to heavy shell deposits among live oysters during immediate premortality and mortality periods. Only on the shells of dead oysters did conchiolin deposits approach the 60-100% level recorded in live oysters during the same period in 1991. Cohort III. which had the highest mortality, also exhibited the lowest prevalence of conchi- olin. only 1 1-17% of the population exhibited light to heavy de- position among live oysters. Histopathological study of tissues from dying cohort I individuals collected on July 20 indicated that 60% had some mantle lesions. Coccoid bodies (Bricelj et al. 1992) were evident in only 15%. and ciliates in another 15%. Cumulative mortalities of oysters held in 6-mm mesh growout trays were half as much as those observed in oysters held in 1-mm mesh trays ( 16 vs. 35%. P = 0.08) (inset. Fig. I). Mortalities for cohort II. held in larger mesh growout trays, ceased to increase 2 wk earlier than mortalities in smaller mesh trays, and final mor- tality was even lower than that of cohort 1 nursery oysters (24%). Anomalous conchiolin deposits were less prevalent in oysters from larger mesh size trays ((3-25%) than in those from smaller mesh size trays (0-40%) before and during mortality. Oyster Growth Patterns All nursery cohorts experienced slower growth rates (slopes of shell height increase between sampling dates) during the period of mortalities (Fig. 2). although apparent growth did not cease en- tirely, probably because smaller oysters died. Maximum shell growth rates of 0.33-0.49 mm d"' were observed among all cohorts before the onset of mortalities. Similar growth patterns were observed during a previous study at the same site (Bricelj et al. 1992). The average shell height of the nursery oysters in the larger mesh trays on September 28 was not significantly greater (P = 0.125) than that of oysters in the smaller mesh trays (41.6 vs. 38.8 mm; inset. Fig. 2). despite the differences in mortalities. Phyloplankton Two microplanktonic species occurred at high densities at the nursery site during 1993 (Fig. 3). A bloom of A/, ndvuin (1.150 50- DATE Figure 2. Mean ( ±SE) shell heights of three cohorts and oyster cohort II grown in 1- and 6-mm mesh (inset) at the FMF Co. oyster nursery during 1993. (A) Cohort I. (D) Cohort II, 1-mm mesh trays. (■) Cohort II, 6-mm mesh trays. (C ) Cohort III. Error bars smaller than the symbol are not indicated. Arrows indicate the onset (closed ar- rows) and end (open arrows) of mortalities of each cohort. Juvenile Oyster Disease 323 cells mL" ', on July 6), a red tide-producing ciliate (ranging in size from 15 to 70 fi-m), was observed before the onset of mor- talities in cohorts I and II. G. sangiiineum was not detected until July 20, 1 wk after oyster mortality began in cohort I, and peaked on August 3 (458 cells mL ~ ' ). at the onset of mortalities in cohort III, and again on August 24 (212 cells mL"'). The first bloom occurred concurrently with the maximum mortality in cohort I oysters. Comparable cell densities of G. sangidneum were ob- served at this study site in 1991 (Bricelj et al. 1992), suggesting that this species may be established in Oyster Bay. The cell con- centrations of G. sangumeum were approximately the same at surface, intermediate, and bottom depths (P = 0.56). Diatom species, known to be of high food quality for bivalves, were abun- dant both before and after the midsummer M. ruhriim and G. sanguiiwum blooms, when oyster growth was also good, but not during blooms. Thalassiosira pseudonana, Skeletonema costatuin. Chaetoceros spp., and Ceratulina bergonii were dominant during early summer, whereas 5. costatum comprised from 65 to 99% of total diatoms in late summer (inset. Fig. 3). Environmental Parameters: Temperature and Salinity Surface water temperature varied between 16.5 and 27.5°C during the study period and exceeded 20°C from mid-June through mid-September (Fig. 4). The outbreak of mortalities occurred dur- ing the period of elevated surface water temperatures (range 22- 27°C) between mid-July and early September. Salinities ranged from 24 to 27 ppt. Salinity was within the optimal range for the growth of most Vibrio spp. (Baumann et al. 1984) and also for the growth ofC. v(>g(>(/ca (Mannet al. 1991 ). The variations recorded in surface water temperature and salinity were typical of previous years at this site. Trends in Vibrio spp. Densities Vibrio spp. concentrations in both surface and bottom waters at the nursery site were low (1 — 25 CFU mL" '; inset. Fig. 5), with the exception of an unusually high peak observed in surface waters on August 3 ( 185 CFU mL " ' ). Reference site measurements ( 1-5 CFU mL" ') were always lower than measurements adjacent to the nursery floats. Vibrio spp. concentrations in sediments at the nurs- ery site fluctuated markedly and were highest (500-1,300 CFU g" ') between June 15 and July 27, before exponential increases of Vibrio spp. in oysters themselves and the onset of mortality were 40 1400- 1200 ■^ 1000 i3 800 _i UJ O 600 400 200 MORTALITY PERIOD -*m » May Jun Jul Aug Sep Oct DATE Figure 3. Cell densities of G. sanguineum (■) and M. rubrum (O) and relative concentrations (inset) of diatoms, dinoflageliates (including G. sanguineum), and other flagellates in surface water at the oyster nurs- ery in Oyster Bay, NY, in 1993. The thick horizontal line indicates the period of juvenile oyster mortalities. 35 30 25 20 15- 10 TEMPERATURE SALINFTY MORTALITY PERIOD 40 35 30 8 25^ Z ■15 10 May Jun Jul Aug Sep Oct DATE Figure 4. Surface water temperature and salinity at the oyster nurs- ery. Oyster Bay, NY, during the 1993 study period. The thick hori- zontal line indicates the period of juvenile oyster mortalities. observed (Fig. 6). By late July and August, sediment Vibrio spp. concentrations had decreased (<200 CFU g" '). whereas they es- calated in oyster tissues during this mortality period. At the ref- erence sites. Vibrio spp. concentrations in the sediment samples were typically two to three orders of magnitude lower ( 1(3-70 CFU g" ') than those in the nursery sediment samples. Vibrio spp. concentrations in nursery oyster tissues always in- creased at least one order of magnitude approximately 2 wk before mortalities were observed (Fig. 6). Concentrations in oysters from cohorts I and II were in the range of 10"" CFU g " ' for the first few weeks after deployment in nursery trays. An exponential increase to 3 X lO"* CFU g" ' occurred in cohort I oysters between June 15 and June 29, 2 wk before mortality was first observed (Fig. 6A). In cohort II, a similar exponential increase to 2 x lO'* CFU g" ' was observed between June 15 and July 6, again 2 wk before the first mortality was observed (Fig. 6B). Vibrio spp. concentrations in cohort III were high (10'' CFU g~') at initial deployment, decreased the following week, and then increased again 1 wk before mortalities began (Fig. 6C). Vibrio spp. concentrations de- clined after mortalitites began and then peaked twice in all cohorts generally a few days after the G. sanguineum blooms (arrows), suggesting a contributing stress effect of the dinoflagellate blooms on the oyster. Variation (i.e., timing of increases and decreases) in 1500 1200 o I = o S a 01 I .', I pi«,n- May Jun Jul Aug Sep Oct _ DATE Figure 5. Mean total Vibrio spp. concentrations in nursery bottom sediment (■) and tray debris (biodeposit) from cohort I nursery oys- ters (C) in CFU g wet weight"' and in nursery (13) and reference water samples (•) (inset) in CFl' mL"' during 1993. Arrows indicate the onset (closed arrow) and end (open arrows) of juvenile oyster mortality periods. 324 Lee et al. TABLE 3. Distribution of Vibrio spp. within cohort III oysters (sampled on October S, 1993). O LJJ > 1- IS ZD o May Jun Jul Aug Sep Oct DATE Figure 6. Mean cumulative oyster mortalities ( — 1 and total Vibrio spp. concentrations in CFU g wet weight ' (■) in three cohorts of juvenile oysters held in outdoor floating trays at the oyster nursery during the 1993 study period. The dotted lines represent estimated total mortality. (A) Cohort I, (B» cohort 11, (C) cohort 111. Arrows indicate dates when peak concentrations of C sanguineum were ob- served. Vibrio spp. concentrations among all three cohorts were overall similar (r^ = 0.40-0.85), indicating that the effects of Vibrio spp. were similar among all three cohorts. However, the oyster mor- tality patterns with respect to onset, duration, and intensity were not completely consistent among cohorts, indicating the possible involvement of other variables in addition to Vibrio spp. Vibrio spp. concentrations in reference site oysters (30-2,500 CFU g ') were two- to three-fold lower than in the nursery oysters at the same sampling date. Vibrio spp. densities in tray debris from nursery oysters were typically higher than those of any other sam- ple, varying from 210 to 140,000 CFU g" ' (e.g., cohort I in Fig. 5). Oysters from the reference site had no such debris. Vibrio spp. concentrations in the tray debris remained relatively high during the mortality period, even while declining in the sediments and remaining low in the water. Examination of the partitioning of Vibrio within oysters col- lected on October 5 , 1 993 , indicated that, per gram of wet weight , the Vibrio recovered appeared to be more or less evenly distributed between meats (20%), pallial fluid (32%), and shell (21%) (Table Mean SE % Sample (CFU/g wet wt) (n = 3) of Whole Whole 2926 527 100 Meat/tluid 1523 317 52 Fluid 944 102 32 Meat only* 579 215 20 Shell 628 80 21 Unaccounted 27** * Meat only = meat and pallial tluid - pallial fluid. ** Unaccounted = whole - (meat + fluid + shell). 3). In this analysis, oysters were sacrificed at the end of the sam- pling period, well after mortalities occurred. However, if the par- titioning observed is representative of the whole sampling period, it would suggest that the pallial fluids may act as a transitional bacterial medium between the surrounding external environment and soft tissues; thus pallial fluid continually exposes oyster tis- sues to the potentially pathogenic bacteria. Potential Sources of Vibrio spp. Vibrio spp. were undetectable in sewage outfall water, indicat- ing that the local waste treatment plant was not a source of con- tamination at the time of analysis (Table 2). High Vibrio spp. counts (>700 CFU mL" ') were found in the Isochr\sis microal- gal cultures usually fed to setting oysters (Table 2). A few of the numerically dominant Vibrio isolates from the flagellate and dia- tom cultures (although total Vibrio spp. concentration was low in the T. weissjlotjii tank) were similar to the isolates collected from the nursery float area (sediment, oyster and water samples) such as Vibrio unguillarum and Vibrio splendidiis . Although this analysis was carried out at a single sampling time, it demonstrated that Vibrio spp. were also present within the hatchery. This finding was consistent with high Vibrio spp. counts from the hatchery larval settling tank (Table 2). Apparently, at the time of sampling, the larvae and postlarvae were being fed contaminated algal cul- tures, suggesting that contamination of oysters with some Vibrio spp. may have occurred even before they were exposed to bay water. Cohort III was present in the hatchery at the time of these samplings. This was consistent with the high Vibrio spp. counts observed in cohort III before they were even transferred to the nursery floats (Fig. 6C). However, the counts represent total Vibrio spp. populations and the actual pathogens may not have been present at all. Oyster Mortalities and Reisolation of Vibrio spp. in Challenge Experiments Average mortality in the first challenge experiment varied from 2 to 6% in controls and in five of the six treatments with bacterial isolates (Fig. 7A). One Vibrio isolate, ST-131, caused signifi- cantly higher mortality than controls ( 15%; F = 5.99; P < 0.001 ) and other treatment groups (F = 4.96; P = 0.002). The second experiment (Fig. 7B) produced much higher mortalities in all groups. Two batches of oysters injected with Vibrio isolates ST-78 and ST-131 showed higher mortalities (73 and 68% at F = 4.75; Juvenile Oyster Disease 325 Q 7 days ■ 14 days ^ 21 days Q 30 days Blank FSW E. coli 78 98C 105C 103B 172 131 ^ tr o m > Z) o 100- 80- 60- 40- 20- 0 B n. m Blank FSW E.coli 78 131 98C 103B 172 68 c - FS ^ ^^ r^ ^ ± - ^i — 1^1 100 80 60 40 20 Blank Notch E.coli 78 131 78+131 74 74+46 46 TREATMENT Figure 7. Mean (±SE) total mortalities of healthy juvenile oysters injected with Vibrio isolates in three experiments. (A) Juvenile oysters obtained from the East Hampton hatchery, NY, notched and injected with filtered seawater, suspensions of E. coli, or suspensions of Vibrio isolates or notched but not injected (blank), incubated for 30 days. (B) Juvenile oysters obtained from Haskin Shelirish Research Laboratory, NJ, subjected to similar experimental treatments except that they were crowded into mesh bags, supplemented with bacteria once every other day, and maintained for 14 d. Filtered seawater control in panel B is suspect because the filter was compromised. (C) Juvenile oysters were obtained from Haskin .Shellfish Research Laboratory. They were crowded into suspended mesh bags in aquaria containing sediment and supplemented with bacteria once every other day; some were treated with a combination of bacterial isolates and maintained for 21 d. Mortalities in all challenge experiments are represented in 7-d in- crements. Probability ranges of significance in difference between con- trols and other groups are indicated by asterisks above the bars. Bars indicate that the sample mean is not significantly different. P < 0.001 and F = 4.75; P «: 0.003, respectively) than those injected with other isolates or control groups. All Vibrio spp. treatments also showed higher mortalities than control groups (F = 7.71; P = 0.016). Examination of shells revealed blister-like organic deposits in the second experiment, including controls in- jected with E. coli. hut they were not the typical JOD deposits observed in field-deployed nursery oysters. Biochemically and physiologically identical Vibrio spp. were reisolated from ST-78 and ST-131 oyster batches at the end of the second experiment. The third experiment (Fig. 7C) showed that a combination of two isolates did not produce significantly higher mortality (P = 0.67) than single-isolate treatments, and there were no significant dif- ferences among treatments (P = 0.2 ~ 0.9), except ST-46, which showed higher mortality (F = 4.96; P = 0.047) than other treat- ments. However, all Vibrio spp. treatments showed higher mor- talities than control batches (F = 4.49; P < 0.001). In the third challenge experiment, most of the notched control oysters showed slight brown depositions around the notched area (Fig. 8A. arrow). Control oysters without notching showed no such depositions. Brown deposition is probably a defense response to the stress of being notched. Not all experimental oysters in- jected with Vibrio isolates displayed symptoms typical of JOD. However, those that did exhibited indications of mantle retraction (Fig. 3B, arrow) and conchiolin deposition (Fig. 8C, arrow). DISCUSSION The Role of N onbacteriological Variables in JOD Temperature and Salinity Salinity and temperature were within the normal range for the study site and for the growth of eastern oysters. In this study, the onset of oyster mortalities followed a period of increasing temper- atures and did not occur until they exceeded 22°C. This finding agrees with previous studies, in which high water temperatures (>20°C) were identified as the most important environmental pa- rameter related to the timing of mortalities, but were nonetheless suggested to play only a secondary role in JOD (Bricelj et al. 1992, Ford 1994). The major consequence of high temperature is its effect on bacterial proliferation, because many Vibrio spp. ac- tively grow and multiply at temperatures >20°C. Phytoplankton Blooms Nightingale (1936) reported an association between mortalities of oysters, Ostrea lurida, especially young stages, and red tides of G. sangumeum in Oakland Bay. WA. In this study, however, the timing of the first G. sanguineiim bloom relative to observed oys- ter mortalities does not support the hypothesis that this dinoflagel- late is the primary etiological agent for JOD. Lewis (1993) deter- mined that JOD was a transmissible disease without the presence of G. sanguineiim. Wikfors and Smolowitz (1994) found that an isolate of G. .sanguineiim from Oyster Bay did not produce JOD symptoms, i.e., abnormal conchiolin deposition in 10-mm juve- nile oysters. However, copious pseudofeces production was ob- served during the first 2 wk of exposure and dinoflagellate cells were fed at a maximum concentration of only four cells per mil- liliter, two orders of magnitude lower than that attained in the field during our study. Thus, blooms of G. sanguineiim may act as a contributing stress factor. It is also noteworthy that peak abun- dance of G. sanguineum and M. rubrum coincided with marked reductions in the concentration of other phytoplankton species, especially diatoms that commonly support good growth of bi- valves. Furthermore, high Vibrio spp. densities in juvenile oysters and water samples from nursery floats coincided with G. san- guineiim blooms. Hence, this species may indirectly affect the 326 Lee et al. Figure 8. Juvenile eastern oysters from the third challenge experiment. (A) Notched control oyster. Note the slight conchiolin deposition around the notched area. (B> Oyster injected with Vibrio isolate (ST-78) exhibiting mantle retraction in the form of irregular depositions. (C) Oyster injected with a combination of two Vibrio isolates (ST-46 + 74) showing early stages of "symptomatic" conchiolin deposition. Arrows indicate deposition. progression of JOD among nursery oysters by influencing bacterial densities in tlie water and in tlie juvenile oysters. An association between phytoplankton or zooplankton blooms and elevated concentrations of Vibrio spp. and other bacteria have been reported in a number of previous studies (Kaneko and Col- well 1973, 1975. 1978. Huq et al. 19X3. Hoppe 1984. Williams and LaRock 1985. Romalde et al. 1990a. 1990b. Montgomery and Kirchman 1993). During this study, a bloom of M. rubrum pre- ceded Vibrio spp. increases in oyster and oyster mortalities. How- ever, a stong association between Mesodinium cell densities and bacterial abundance in the water column, as observed by Crawford et al. (1993). was not apparent in this study. Furthermore. M. rubrum has not been reported to be toxic and is consumed by a wide range of organisms, from mysids to oysters (Lindholm 1985). Oyster Age and Size Mortality occurred among oysters within the susceptible size range of 10-16 mm. When mortality was first observed in the nursery floats, the mean size of oysters was —20, ~16, and —12 mm for cohorts I, II. and 111, respectively. Mortalities of cohort I oysters did not occur until almost 9 wk after deployment, whereas deaths in the other cohorts took only 5-7 wk to appear (Table 1). Cohort 1 may have taken longer than the others to accumulate a critical level of pathogens because pathogen concentrations were relatively low in the plankton and/or because pathogen prolifera- tion in oysters and oysters' potential filtration/pumping activity were reduced at lower temperatures. By the time additional deter- minants, e.g., higher temperatures and deleterious plankton blooms (stress), were present in the water column, cohort I oysters had presumably grown large enough to be less susceptible to these factors, as reflected in this cohort's lower overall mortality. The opposite may be true for the oysters deployed later. From an energetics standpoint, small individuals, which have a higher met- abolic rate per unit body weight and relatively lower energy stores, are more likely to exhibit nutritional stress than larger individuals (Galtsoff 1964, Fisher 1988). Oysters spawned and deployed later in the growout season hence smaller individuals are thus more susceptible to JOD. Vibrio spp. Concentrations in Relation to Oyster Mortalities The initial exponential increases in Vibrio spp. concentrations were observed in all three cohorts before oysters began dying. Thus, bacterial proliferation was not the result of dead and dying oysters providing a suitable medium for Vibrio spp. growth. This observation provided the first evidence that Vibrio spp. may be involved in JOD. Before this exponential growth, relatively stable Vibrio spp. levels in all three cohorts were probably maintained by a dynamic equilibrium between filtration and egestion or destruc- tion of Vibrio spp. Because of Vibrio spp. proliferation m the oysters' immediate environment, e.g.. debris in tray and in/on the oysters themselves, stressed, crowded juveniles may eventually become vulnerable to disease. Quantitative bacterial analysis con- ducted in other studies has found high counts of Vibrio spp. in both "sick" oyster and clam intervalval fluids, ranging between 10'' and lO*" cells mL~' (Lovelace et al. 1968, Maes and Paillard 1992). Similar concentrations of Vibrio spp. were found in this study in the pallial fluid between mantle and shell. Although not investigated, total Vibrio spp. increases in oysters during this study probably included the amplification of pathogenic Vibrio spp. The critical concentration of pathogenic Vibrio spp. can be different for each cohort because of the varying conditions of juvenile oysters and their environment. It appears that the intensity and timing of mortality are also influenced by oyster size and condition (i.e., prior history) at the time of infection. All other Vibrio spp. increases after the initial premortal ity rises were prob- ably due to active bacterial proliferation inside already infected, dying, or compromised juvenile oysters or their immediate envi- ronment. Mortality can be minimized by growout methods that may pre- vent a critical level of potentially pathogenic Vibrio spp. or other possible agents from accumulating around the oysters. Mortalities associated with JOD have been shown to be lower among oysters placed on a less densely packed raft (Bricelj et al. 1992). In this Juvenile Oyster Disease 327 study, oysters placed on rafts with u larger mesh size also showed lower mortalities. This phenomenon may be attributed to the di- lution of bacteria caused by improved water exchange around the oysters or to the improved condition of oysters themselves reared under these practices. Another possible point of infection not re- lated to the "'contaminated environment" of the nursery site may have occurred during larval rearing and settlement inside the hatchery. This hypothesis is supported by high Vibrio spp. con- centrations in microalgal cultures and hatchery larval and postset tank waters even before oysters (cohort III) were deployed to the raft system (Table 2). It must be emphasized, however, that the correspondence be- tween total concentrations of all Vibrio spp. and mortality in oys- ters within a given cohort does not establish causality. Many bac- teria may simply represent background nonpathogenic popula- tions, or opportunistic infections of compromised oysters, rather than pathogenic Vibrio spp. In this study, the most often identified species of Vibrio were Vibrio parahciemoiyticus, V. angnilldriim. Vibrio vulnificus, and V. splendidus — all known pathogens to varying degrees. However, concentrations of individual Vibrio species are unknown, and the frequency of the occurrence of spe- cific bacteria was not determined. The examination of temporal patterns in Vibrio spp. concen- trations, especially before the start of mortalities, gives some in- sight into the Vibrio-IOD relationship. First. Vibrio spp. concen- trations (1-185 CFU niL" ') observed in the water column showed no relationship to oyster mortality patterns. Nevertheless. Vibrio spp. were present at all times, although at low concentrations. There was a significant correlation between Vibrio spp. concen- trations in surface water and sediment before the onset of oyster mortalities (r^ = 0.869), suggesting that the bay bottom may have acted as a source of Vibrio spp. to the oysters. An observed '■pulse'" of Vibrio spp. in the water samples is not necessary in order to postulate a transfer from sediment to oyster because of the large dilution effect of waters overlying the sediments and the high pumping rate of oysters. These observations do not. however, clearly demonstrate whether sediment was a source for Vibrio spp. found at the nursery site. The FMF hatchery uses the same nursery location year after year. This presumably increases the inoculum of bacteria and or- ganic nutrient loading to the nursery sediment compared with the reference site, through dead and dying oysters, debris, sediment- ing planktonic blooms, and other aquatic animals attracted to the float area. All of these can act as a source for reinfection by providing a medium for bacterial proliferation and by sediment resuspension (Kaneko and Colwell 1973. 1978. Ruby and Morin 1979. Williams and LaRock 1985, Enger et al. 1989. Buck 1990). Under stressful conditions, e.g.. low nutrients, extreme tempera- tures, crowding. Vibrio spp. can develop into viable but noncul- turable resting stages, reinoculating the water column when envi- ronmental conditions improve (Parsons et al. 1984, Molitoris et al. 1985, Williams and LaRock 1985, Pathck et al. 1988). During this study. Vibrio spp. counts in the sediments declined late in the summer season, even though temperature and probably nutrient levels remained favorable for bacteria including Vibrio spp. One of the factors contributing to this decrease could be competition with other bacterial populations in the sediment (Hood and Ness 1982). Enough Vilnio spp. may have survived through the winter to grow and reinoculate the water and oyster during the next sea- son. It appears that the sediment could act as a source of Vibrio spp. early in the growout season and become less important as the season progresses, especially after the oysters and tray debris are inoculated. Debris immediately surrounding the oyster most likely provides a habitat that promotes the growth of Vibrio spp. and other bacteria. Thus, growout tray debris is likely to be an impor- tant transient reservoir for potentially pathogenic Vibrio spp. that promotes additional hostile conditions for the oysters by creating anoxic microenvironments as oysters die and decay. Vibrio spp. as Potential Pathogens Challenge experiments showed that isolates phenotypically similar to V. anguillarum (ST-13I) and Vibrio alginoiyticus (ST- 7-8) caused high juvenile oyster mortalities. Although there was only slight evidence of JOD-like anomalous conchiolin shell de- posits in some of the experimental oysters examined, the high mortalities caused by injected Vibrio isolates strongly suggest that these bacteria are pathogenic under certain conditions. These con- ditions include high water temperatures, crowding, and high con- centrations of specific bacteria. Thus, much higher mortalities were observed in the second and third challenge experiments, in which these conditions were amplified. The paucity of typical JOD symptoms among experimentally infected oysters is a liability in assigning a species of Vibrio as an agent of JOD and thus in fulfilling Koch's postulates. Such a discrepancy does not. how- ever, invalidate the hypothesis. Environmental conditions imposed on oysters during the challenge experiments slightly misrepre- sented in situ nursery conditions. Oysters may have been stressed differently, so that symptomatic conchiolin deposits were not properly induced before mortality ("acute" JOD). The somewhat older and larger oysters available for use in the challenge experi- ments may not have responded in the same way as smaller juvenile oysters afflicted with JOD. The typical conchiolin deposit may be a defense mechanism of juvenile oysters that are in a ""high- growth" mode. Larger oysters, as well as oysters in poor condi- tion, may experience mortality without showing JOD signs (Lewis 1993). Alternatively, the trauma of injections and containment in a closed system may have caused oyster death to take place through "acute toxin-mediated infection"' before oysters could use their defense systems to build conchiolin deposits (Elston et al, 1982, Elston 1984. Paillard and Maes 1994, Paillard et al. 1994). Another consideration is the fact that only nine isolates among 200 tentatively identified bacteria were used in these experiments. It is thus possible that the causative agent(s) of JOD may not have been among the selected nine isolates. V. anguillarum and V. alginoiyticus. both of which induced high mortalities in the challenge experiments, have been isolated from coastal environments (Tubiash et al. 1973, Sindermann I977,DiSalvoetal. 1978, Sakazaki and Balows I98I,Kent 1982, Larsen et al. 1988, DiSalvo 1995). Their involvement as bivalve pathogens or in producing substances toxic to various bivalves, including oysters, is frequently reported (Tubiash et al. 1970, 1973. Sindermann 1977. 1990. Nottage and Birkbeck 1986. 1987. Birkbeck et al. 1987). but JOD etiology has not been previously described for these species. It is feasible that new pathogenic strains of these bacteria that produce toxins or other imtants (Kaper et al. 1979. Colwell 1984. Nottage and Birkbeck 1986, 1987) may have established themselves at this particular site. Attempts to elucidate causes for mortalities in aquaculture of- ten focus on a search for primary pathogens, when the real cause may be environmental, nutritional, or physiological and may in- 328 Lee et al. volve facultative, opportunistic microorganisms acting on com- promised hosts. Factors such as water quality and other environ- mental variables and the physiological status of the animals must also be considered. In order to discover which species of Vibrio. if any, is the true etiological agent of JOD, additional challenge experiments should be performed with other Vibrio and non-Vibrio isolates to repeatedly and fully satisfy Koch's postulate. These experiments should entail varying environmental factors in an at- tempt to better simulate conditions found at the nursery site. Fac- tors such as the continuous rather than batch delivery of food and bacterial sources and the inoculation of susceptible oysters with combinations of different bacterial isolates in the presence and absence of high plankton densities should be incorporated into future experimental designs. Oyster strains resistant to pathogenic Vibrio spp. (Farley et al. 1995) and conditions favoring JOD also need to be identified to provide aquaculturists with additional means for reducing losses due to JOD. Although not yet conclu- sively demonstrated to be the definitive pathogen, species of Vibrio appear to have a strong involvement in JOD development, as suggested by field observations and challenge experiments, and still must be considered to be a potential candidate for the etio- logical agent of JOD. ACKNOWLEDGMENTS We thank our colleagues for frank and stimulating discussions during the Annual Aquculture Seminar workshops on JOD held at the NMFS Laboratory. Milford, CT, as well as personnel at Frank M. Flower and Sons Inc.. especially D. Rclyea, D. Berg, and J. Zahtila, for their cooperation and participation in field sampling. We also thank E. Carpenter (MSRC) for phytoplankton species identification. Bob Barber for histopathology and hatchery man- agers, G. Debross, (Haskin Shellfish Research Laboratory), and J. Aldred (Town of East Hampton) for providing us with oysters used in challenge experiments. Any opinions or conclusions expressed in this article are those of the authors and do not necessarily reflect the views of the Northeastern Regional Aquaculture Center (NRAC). This project was supported, in part, by NRAC subcon- tract number 555344 and grant numbers 92-38500-7142 and 90- 38500-5211 Amencan Public Health Association (A.P.H.A.). 1970. Recommended Procedures for the Examination of Sea Water and Shellfish. 4th ed. APHA, New York, pp. 1-99. Anderson, D. A. & R. J. Sobieski. 1980. Introduction to Microbiology. 2nd ed. C. V. Mosby, St. Louis, pp. 2-152, Balows, A. 1974. Current Techniques For Antibiotic Susceptibility Test- ing. C. C. Thomas. Spnngfield, pp. 1-173. Baumann, P., A. L. Fumiss & J. V. Lee. 1984. Genus I Vibrio, pp. 518-538. In: N. P. Krieg and J. G. Holt (eds.). Sergey's Manual of Systematic Bacteriology, vol. 1. Williams & Wilkins, Baltimore. Biolog GN Microplate Instruction for Use. 1992. Biolog Inc., Hayward, California, pp. 1-12. BioMerioux Vitek, Inc. API 20E Instruction for Use 1992. Hazelwood, Missouri, pp. 1-8. Birkbeck, T. H.. J. G. McHenry & A. S. Nottage. 1987. Inhibition of filtration in bivalves by marine Vibrios. Mar. Biol. 95:247-248. Bricelj, V. M., S. E. Ford. F. J. Borrero, F, O. Perkins, G. Rivara. R. E. Hillman, R. A. Elston & J. Chang. 1992. Unexplained mortalities of hatchery-reared, juvenile oysters, Crassostrea virgiiiica (Gmlin). J. Shellfish Res. 11:331-347. Brock, T. D. & M. T. Madigan, 199L Laboratory Manual. Biology of Microorganisms. 6th ed. Prentice Hill. New Jersey, pp. 172-229. Brock, T. D., M. T. Madigan, J. M. Martinkod & J. Parker. 1994. Bi- ology of Microorganisms. 7th ed. Prentice-Hill, New Jersey, pp. 356- 531. Bryant, T. N., J. V. Lee, P. A. West & R. R Colwell. 1986. A proba- bility matrix for the identification of species Vibrio and related genera. J. Appl. Bacterial. 61:469-480. Buck, J. D. 1990. Isolation of Candida albicans and halophilic Vibrio spp. from aquatic birds in Connecticut and Flonda. Appl. Environ. Micro- biol. 56:826-828. tiurdon, K. L. & R. P. Williams. 1968. Microbiology. 6th ed. The Mac- Millan Company. New York, pp. 329-330. Colwell, R. R. 1984. Vibrios in the Environment. Wiley cS: Sons. New York. pp. 1-516. LITERATURE CITED Cowen, S, T. 1974. Cowen and Steel's Manual for the Identification of Medical Bacteria. 2nd ed. Cambridge University Press, New York. pp. 95-180 Crawford, D. W,, L. E. Hawkins. S. Hutchinsons, E. E. Antai, D. A. Purdie & A. P. M. Lockwood. 1993. Red tides of Mesodiniwn rii- brum: evidence for remotely imposed stress on the oyster Osnea edu- lis. Dev. Mar. Biol. 3:389-394. Difco Dispens-O-Disc for Susceptibility Testing Instruction. 1991. Difco Laboratones. Detroit, pp 1— I. DiSalvo. L. H. 1995. Chronic infection of scallop broodslock. lOth Inter- national Pectinid Workshop. April 27-May 2, 1995. Cor. Ireland, pp, 93-94 (abstract). DiSalvo, L. H., J. Black & R, Zebal. 1978. Vibrw angiiillanim and larval mortality in a California coastal shellfish hatchery. Appl. Environ. Microbiol. 35:219-221. Elston, R. A. 1984. Prevention and management of infectious diseases in intensive molluscs husbandry. J. World Mariculnire Soc. 15:284-300. Elston, R. A.. E. Elliot & R. R. Colwell. 1982. Conchiolin infection and surface coating Vibrio: shell fragility, growth depression and mortali- ties in cultured oysters and clams, Crassostrea virginica. Ostrea edulis and Mercenaria mercenaria. J. Fish Dis. 5:265-284. Enger, O., B. Husevag & J. Goksoyr. 1989. Presence of the fish pathogen Vibrio salmonicida in fish farm sediments. Appl. Environ. Microbiol. 55:2815-2818. Farley, C. A.&E. J. Lewis. 1993. Juvenile oyster mortality studies-1992: histopathology. pathology, epizootiology. J. Shellfish Res. 12:107 (ab- stract). Farley, C. A., F. G. Kern, E. J. Lewis & W. Blogoslowaski 1992 In- fectious agent, Ulceraiiis spraguei g.n,. s.p.n. (Protista Incertae se- dis). associated with fatal disease in juvenile oysters. Crassostrea vir- ginica. in New York water. 12th Annual Shellfish Biology Seminar. Feb, 24-26, 1992, Milford, Connecticut. 11 pp. (abstract). Farley, C. A., E. J. Lewis, D. Relyea & J. Zahtila. 1995. Studies of resistance in progeny of brood stock selected from juvenile oyster disease (JOD) survivors. J. Shellfish Res. 14:242 (abstract). Juvenile Oyster Disease 329 Fisher, W. S. 1988. Environmental influence on host response; environ- mental mfluence on bivalve hemocyte function. 4m. Fisheries Soc. Sp. Piihl. 18:22?-237. Ford, S. E. 1994. Unexplained Mortalities of Hatchery-Reared Oysters. Crassoslrea virginica. in the Northeast. A Follow-Up Study, Apnl 1, 1992 to March 31, 1993. Final Technical Report to the Northeastern Regional Aquaculture Center (NRAC). University of Massachusetts at North Dartmouth, Massachusetts, pp. 1-40. Ford, S. E & C. Paillard. 1994. A comparison of juvenile oyster disease in the USA and brown nng disease of manila clams in Europe. Tech- nical papers. J Shellfish Res. 13:314 (abstract). Galtsoff, P. S. 1964. The American oyster Crassoslrea virginica Gmlin. U.S. Depanment of the Interior, Fish. Bull, of the Fish and Wildlife Service 64, 457 pp. Glantz, S. A. 1992. Primer of Biostatistics. 3rd ed. McGrow Hill, New York. pp. 35-376. Hillman. R. E. & R, A. Elston. 1992. A study of juvenile oyster mortal- ities at the Frank M. Flower and Sons Oyster Hatchery. 12th Annual Shellfish Biology Seminar. Feb. 24-26. 1991. NMFS Laboratory. Mil- ford, Connecticut II pp. (abstract). Holt. J. 0., N. R. Kneg, P. H. A. Sneath, J. T. Staley & S. T. Williams. 1994. Sergey's Manual of Determinative Bacteriology. 9lh ed. Williams & Wilkins, Baltimore, pp. 190-274. Hood, M. A. & G. E. Ness. 1982. Survival of Vibrio cholerae in estuarine waters and sediments. Appl. Environ. Microbiol. 43:578-584. Hoppe, H. G. 1984. Attachment of bacteria: advantage or disadvantage for survival in the aquatic environment, pp. 283-301. In: K. C. Marshall (ed.). Microbial Adhesion and Aggregation. Springer- Verlag, New York. Huq, A., E. B. Small, P. A. West, M. 1. Huq, R. Rahman & R. R Colwell. 1983. Ecological relationships between Vibrio cholerae and planktonic crustacean copepods. Appl. Environ. Microbiol. 45:275- 283. Kaneko, T. & R. R. Colwell. 1973. Ecology of Vibrio parahaemolyliciis in Chesapeake Bay. J. Bacterial. 113:24-32. Kaneko, T. & R. R. Colwell. 1975. Adsorption of Vibrio parahaemolyti- cus onto chitin and copepods. Appl. Microbiol. 29:269-274. Kaneko, T. & R. R. Colwell. 1978. The annual cycle of Vibrio parahae- molyticus in Chesapeake Bay. Microb. Ecol. 4:135-155. Kaper, J. B., H. Lockman, R. R. Colwell & S. W. Joseph. 1979. Ecol- ogy, serology and enterotoxm production of Vibrio cholerae in Ches- apeake Bay. Appl. Environ. Microbiol. 37:91-103. Kent, M. L. 1982. Charactenstics and identification of Pasleurella and Vibrio species pathogenic to fishes using API-20E (Anaiytab Products) multitube test stops. Can. J. Fish. Aquai. Sci. 39:1725-1729. Larsen, J. L., H. B. Rasmussen & I. Dalsgaard. 1988. Study of Vibrio anguillarum strains from different sources with emphasis on ecological and pathobiological properties. Appl. Environ. Microbiol. 54:2264- 2267. Lee, J. V. 1979. Identification o( Aeromonas. Vibrio and related organ- isms, pp. 152-166. In: F. A. Skinner &D. W. Lovelock (eds.). Iden- tification Methods for Microbiologists. 2nd ed. Academic Press, Or- lando. Lewis, E. J. 1993. Transmission Studies of a Disease Affecting Juvenile Oyster, Crassoslrea virginica Gmelin, in the Northeastern United States. M.S. Thesis. University of Maryland at Baltimore. Marine Estuarine Environmental Science, Maryland, pp. 2-105. Lewis. E. J. & C. A. Farley. 1994. Effects of salinity and selected treat- ments on juvenile oyster disease. 14th Annual Aquaculture Seminar, Feb. 22-24, 1994, Milford, Connecticut. J. Shellfish Res. 13:316 (ab- stract). Lindholm, T. 1985. Mesodinium rubrum — a unique photosynthetic ciliate. Adv. Aqual. Microbiol. 3:1^8. Lovelace, T. E., H. Tubiash & R. R. Colwell. 1968. Quantitative and qualitative commensal bactenal flora of Crassoslrea virginica in Ches- apeake Bay. Proc. Natl. Shellfish. Assoc. 58:82-87. Maes, P. & C. Paillard, 1992. Effect du Vibrio PI, pathogene de Riidi- tapes philippinarum. Sur d'aulres especes de bivalves. Les mollusques marins. biologic et aquaculture. IFREMER. Acles de Colloqiies. 14: 141-148. Mann, R., E. M. Burreson & P. K. Baker. 1991. The decline of the Virginia oyster fishery in Chesapeake Bay; Considerations for intro- duction of a non-endemic species. Crassoslrea gigas (Thunberg, 1793). J. Shellfish Res. 10:379-388. Molitoris, E., S. W. Joseph. M. I. Krichevsky. W. Sindhuhaardja & R. R. Colwell. 1985. Characterization and distribution of Vibrio alginolyticus and Vibrio parahaemolyticus isolated in Indonesia. Appl. Environ. Microbiol. 51:1388-1394. Montgomery, M. T. & D. L. Kirchman. 1993. Role of chitin-binding proteins in the specific attachment of the marine bacterium Vibrio har\'eyi to chitin. Appl. Environ. Microbiol. 59:373-379. Nightingale, W. H. 1936. Red Water Organisms: Their Occurrence and Influence Upon Marine Aquatic Animals, With Special Reference to Shellfish in Waters of the Pacific Coast. Argus Press, Seattle, Wash- ington. 24 pp. Nottage, A. S & T. H. Birkbeck. 1986. Toxicity to marine bivalves of culture supernatant fluids of the bivalve-pathogenic Vibrio strain NCMB 1338 and other manne Vibrios. J. Fish Dis. 9:249-256. Nottage, A. S. & T. H. Birkbeck. 1987. The role of toxins in Vibrio infections of bivalve molluscs. Aquaculture 67:244-246. Paillard. C. & P. Maes. 1990. Etiologie de la maladie de I'anneau brun chez Tapes philippinarum: pathogenicite d'un Vibrio sp. C R Acad. Sci Paris. Ser. HI. 310:15-20. Paillard, C. & P. Maes. 1994. Brown ring disease syndrome in the Manila Clam, Ruditapes philippinarum: establishment of a classification sys- tem. Dis. Aquat. Org. 19:137-146. Paillard, C P. Maes & R. Oubella. 1994 Brown ring disease in clams. Annii. Rev. Fish Dis. 4: (in press). Palmer. R. E. 1980. Observation on shell deformities, ultrastructure and increment formation in the bay scallop Argopeclen irradians. Mar. Biol. 58:15-23. Parsons. T. R., M. Takahashi & B. Hargrave. 1984. Biological Oceano- graphic Processes. 3rd ed. Pergamon Press, New York. pp. 169-256. Pathck, S. P., J. W. Bhattacherjee, N. Kaira & S. Chandra. 1988. Sea- sonal distribution of Aeromonas hydrophila in nver water and isolation from river. J. Appl. Bacterial. 65:347-352. Relyea, D. 1992. Unexplained mortalities in juvenile hatchery raised oys- ters. The 12th Annual Shellfish Biology Seminar. Feb. 24-26. 1992. NMFS Laboratory. Milford. Connecticut. 10 pp. (abstract). Relyea. D. 1994. Modifying oyster hatchery management to avoid losses due to unexplained mortality of juvenile oysters. The 14th Milford Aquaculture Seminar. Feb. 22-24. 1994. NMFS Laboratory. Milford. Connecticut. 17 pp. (abstract). Romalde. J. L.. J. L. Barja & A. E Toranzo. 1990a. Vibrios associated with red tides caused by Mesodinium rubrum. Appl. Environ. Micro- biol. 56:3615-3619. Romalde, J. L.. A. E. Toranzo & J. L Barja 1990b. Changes in bactenal populations during red tides caused by Mesodinium rubrum and G\m- nodinnim catenatum in northwest cast of Spain. J. Appl. Bacleriol. 68:123-132. Ruby. E. G. & J. G. Morin. 1979. Luminous enteric bactena of marine fishes: a study of their distnbution. densities and dispersion. Appl. Environ. Microbiol. 38:406-411. Sakazaki, R. & A. Balows. 1981. The genera Vibrio. Plesiomonas, and Aeromonas. pp 1272-1301. In: M. P. Starr, H. Stolp, H. G. Truper, A. Balows & H. G. Schlegel (eds.). The Prokaryotes. Spnngler- Verlag, New York. Sindermann, C. J. 1977. Disease Diagnosis and Control in North .Amer- ican Marine Aquaculture. Elsevier Scientific, New York. pp. 184-225. Sindermann, C. J. 1990. Pnncipal Diseases of Marine Fish and Shellfish. 2nd ed. vols. I & 2. Academic Press, New York, pp. 1-516. Sokal, R. R. & F. J. Rohlf 1981. Biometry. 2nd id. W. H. Freeman & 330 Lee et al. Co.. San Francisco, pp. 345-859. H. Hashimoto. 1989. Occurrence and distribution of Vibrio spp., Lis- Tubiash, H. S., P. E. Chanley & E. Leifson. 1965. Bacillary necrosis, a timella spp.. and Clostridium boiuliium in the Seto Inland Sea of disease of larval and |uvenile bivalve mollusks. J. Bacterial. 90:1036- Japan. Appl. Environ. Microbiol. 55:559-567. 1044, Wikfors. F. H. & R. M. Smolowitz. 1994. A GYmnociinium isolated from Tubiash, H. S., R. R. Colwell & R. Sakazaki. 1970. Marine Vibrios the F. M. Flower Hatcher>' causes no apparent ill effect in juvenile associated with bacillary necrosis, a disease of larval and juvenile oysters. The 14th Milford Aquaculture Seminar, Feb. 22-24, 1994. bivalve mollusks. J. Bacterial. 103:272-273. Milford. Connecticut. J. Shellfish Res. 13:322 (abstract). Tubiash. H. S.. S. V. Otto & R. High. 1973. Cardiac edema associated Williams. L. A. & P. A. LaRock. 1985. Temporal occurrence of Vibrio with Vibrio angnillaritm in the American oyster. Proc Natl. Shellfish species and Aeromanas hydrophila in estuarine sediments. Appl. En- Assac. 63:39-42. viron. Microbiol. 50:1490-1495. Venkateswaran. K . H. Nakano. T, Okabe, K Takayama, O. Matsuda & Jounuil of Shellfish Research. Vol. 15. No. 2. 331-339, 19%. BIOMASS AND DISTRIBUTION OF FIVE SPECIES OF SURF CLAM OFF AN EXPOSED WEST COAST NORTH ISLAND BEACH, NEW ZEALAND MALCOLM HADDON,' TREVOR J. WILLIS.^ ROBERT G. WEAR,' AND VICTOR C. ANDERLINI^ institute of Marine Ecology School of Biological Sciences (Al 1 ) University of Sydney Sydney. NSW 2006. Australia ^Leigh Marine Laboratory: University of Auckland P.O. Box 349 Warkworth. New Zealand ^Island Bay Marine Laboratory 396-402 The Esplanade Island Bay . Wellington, New Zealand ABSTRACT In 1942. Ihe total biomass of five species of surf clam, between depths of 1 and 9 metres below chart datum, along 27.5 km of west coast North Island. New Zealand, open sandy coast was conservatively estimated as being 1.032 t. Approximately 50% was Dosinici anus. 197( was Maeira ctisears. 16% was Paphies donacina, 14% was Mactra murchisoni, and 3% was Spisula aequilalera. This biomass equates to approximately 37.5 t km~ ' of coastline, and the potential harvest from this standing crop is correspondingly limited. The only other common macrofaunal species taken was a sand dollar, Fellaster zelandiae. The numerical density of each surf clam species varied with water depth, each having an optimal range. A trend of increasing density along the surveyed coast (increasing with latitude) was found with F. zelandiae. which suggests that recruitment processes are not always uniform or random along the coast at the scale of this study. A less marked trend was also found with P. donacina. No trends of density with latitude were found with D. anus, M . discors. M . murchisoni, or S. aequilalera. Good recruitment was evident from length frequency distnbutions in all species except P. donacina. It is suggested that the smaller size classes of this species live in shallower water than that sampled. Small S. aeijuilaiera were more common in shallow . inshore waters than offshore, whereas the reverse pattern of size with depth was exhibited by the two Maclru species and by D. anus. P donacina showed no trend of mean length with depth. All species exhibited highly positively skewed histograms of frequency of abundance within dredge tows, indicating that they all had highly aggregated distributions. The relative "patchiness" or degree of clumping for each species appeared to be positively related to its the relative abundance. Target fishing for single species is unlikely to be possible on a commercial basis owing to the high degree of overlap in their distributions and the high levels of patchiness for all species. The facts that the five species have different productivity levels and a wide range of growth rates, combined with the inability to target for particular species, have implications for the management of the resource. The sand dollar, F. zelandiae. can be so abundant that in some areas it could clog fishing gear during prolonged tows and become a severe nuisance to commercial fishing. Any fishery based on this coast would best be managed as a mixed species fishery, and the vanability in available biomass suggests that a constant fishing mortality harvesting strategy (such as current annual yield) would be more appropriate than a constant catch strategy (such as maximum constant yield). KEY WORDS: Surt clams, biomass survey. Mactra, Spisula. Paphies. Dosinia. Fellaster INTRODUCTION locally abundant at other localities around the country (Cranfield et al. 1994). Other species occurring on open coasts include the now The biological communities present in the surf zone off ex- uncommon toheroa, Paphies ventricosa, which usually occurs in- posed sandy shores in New Zealand are poorly known (Morton and tertidally, and the grey tuatua, Paphies suhtriangulata. The latter Miller 1973). Until recently, in New Zealand, the difficulties of is the principal species now taken by recreational harvesters and is working in such an environment have precluded anything other common close inshore in water shallower than 1.0 m below chart than cursory investigations. Occasional storm strandings (Powell datum. A small commercial fishery for P. suhtriangulata, using 19791 and analysis of Maori middens (Carkeek 1966, Butts 1982a. hand picking, is already established in northern New Zealand. In Butts 1982b) indicated that considerable populations of relatively the mid- to late 1980s, interest developed concerning whether any large bivalve molluscs lived subtidally along these shores. How- surt" clam species could form the basis of a commercial dredge ever, in most areas, little more was known beyond the mere pres- fishery (Michael et al. 1990). as with similar species in Italy, ence of these "'surt' clams." North America, and other countries (Caddy 1989). Investigations In New Zealand, the term "surf clam" refers to a number of were therefore begun to determine the distribution and abundance bivalve mollusc species. The five main surf clams on the Well- of each species and methods of harvesting them efficiently under ington West Coast are the yellow tuatua Paphies donacina (Mc- New Zealand conditions. sodesmatidae). the venus clam Dosinia anus (Veneridae). two A number of studies had already been carried out by the New similar trough shells Mactra discors and Mactra murchisoni, and Zealand Ministry of Agriculture and Fisheries along with the New the triangle clam Spisula aequilalera (Mactridae). Two other spe- Zealand Fishing Industry Board staff. A preliminary investigation cies. Bassina yalei and Dosinia subrosea also occur and can be testing two different hydraulic dredges, concentrating primarily on 331 332 Haddon et al. a dense bed of tuatua. P . donacina. was earned out at Peka-peka (Fig. 1) on the Wellington West Coast (Michael et al. 1990). The authors did not produce a biomass estimate from these preliminar>' trials, but the high catch rates reported (an average of approxi- mately 58 animals m " *) may have fueled hopes for the existence of a large virgin resource along New Zealand oceanic beaches. Surf clam resources in Cloudy Bay (Cranfield and Michael 1987) have also been investigated (Fig. 1); however, the survey design used was neither stratified nor random and the biomass estimate produced was acknowledged to be "at best very approximate." A detailed tag/recapture study of growth rates has also been carried out successfully in central New Zealand, providing information on the relative productivity of the main species (Cranfield et al. 1993). Exploratory fishing to establish the presence of harvestable concentrations of surf clams, over a broader geographic scale, has been carried out around New Zealand (Cranfield et al. 1994). In that study, highly detailed, small-scale, stratified random surveys were made of the surf clams found along 450-m-wide strips of coast between 1 and 9 m below chart datum at 16 sites around the New Zealand coast. The results of three such detailed site surveys on the Welling- ton west coast suggested that the resource could possibly sustain a viable commercial fishery (Cranfield et al. 1994). However, the biomass estimates were originally made with an assumed dredge efficiency of 65%, determined in a previous survey (Michael et al. 1990). If these estimates are extrapolated to tonnes (t) per kilo- meter of coast (to 9-m depth) and are adjusted for 100% dredge efficiency, they become 25, 95.8, and 53.2 t. respectively, or an average of 57.98 t km~ ' . However, because extrapolation beyond the extent of the three narrow surveyed areas would be invalid, the true size of the resource remained unknown. This study was based on a stratified random survey of biomass and population sizes conducted along a 27.5-km stretch of coast and was designed to facilitate the setting of total allowable catches for the five main surf clam species in that area. Potential recruitment was investigated by determining the population size structure of each species, as indicated by length frequencies. The areal distribution of the five surf clam species was also investi- gated to determine the degree of patchiness in their distributions, as was the effects of depth on distribution, and to determine whether their distributions would allow each species to be indi- vidually targeted. Thus, for each species, the survey samples were also used to investigate how both clam density (biomass) and shell length related to depth. The patchiness of each species was exam- ined first by consideration of how density varied relative to geo- graphical position, and second, by graphical examination of the relative frequency of different levels of abundance. METHODS General Methods Figure 1. Location map showing the general position of the study area in relation to the rest of New Zealand and the detail of the coast between the Manawatu River and the Waitohu Stream. This study extended over a distance of approximately 27.5 km of coast between the Manawatu River and Waitohu Stream (Fig. 1 ) along the Wellington west coast. Sampling began 1 .0 km south of the Manawatu River because of dangerous fishing conditions near the river mouth. The analyses were restricted to the five species in the area sufficiently abundant to have commercial potential; D. emus. P. donacina, M. murchisoni , M. discors. and S. aequilat- era. The echinoderm Fellaster zelandiae, a sand dollar, was also considered because it proved to be the most numerically abundant macroorganism occurring in the samples, with the highest biomass in some areas. Survey Design A stratified random sampling design was adopted for this study. Because surf clams are not evenly distributed relative to depth (Anderlini and Wear 1991, Cranfield and Michael 1992), the survey area was stratified by depth. Before any dredge sam- pling, a bathymetric survey was made of the coastline between the Manawatu River and the Waitohu Stream from 1 .0 and 9.0 m chart datum (Fig. 1). This was done from a 6.0-m boat with a depth sounder (accurate to 0.2 m) and a Magellan PRO 1000 GPS re- ceiver to determine geographical position. The depth strata pro- duced by this method were only approximate indicators of depth. All depths are described as metres below chart datum. The four depth ranges selected were 1-3 m, 3-5 m, 5-7 m. and 7-9 m. The depth range of each stratum was 2.0 m. which, on that coast, meant that each stratum would have a minimum width of between 100 and 200 m. This was necessary in order that dredge tows, which, because of the influence of wind and/or current, were rarely parallel with the coast, would fit comfortably within a single stratum. The strata limits were interpolated from the map of depth soundings. Although they all tended to be parallel to the shore, there were some local deviations both inshore and offshore, par- ticularly near river and stream outlets. The survey area was divided into geographical sectors along the coast in order that each stratum was no more than approxi- mately 3 km". The sector boundaries were based on the location of river and stream mouths because such boundaries are accepted by BioMASS AND Distribution of Exposed Coast Surf Clams 333 the Maori people as legitimate boundaries in matters relating to land claims and fishing rights. This division produced five geo- graphical sectors along the coast, each with four depth strata, providing a total of 20 strata. The stratum areas varied between approximately 0.9 and 3.1 km" with an average area of approxi- mately 2.12 km", whereas the total survey area, between 1.0 and 9.0 m. was approximately 38.5 km". A total of 127 samples were taken, each of 50-m tow length. Initially, 100 random stations were allocated to the 20 strata, ap- portioned according to their relative area. A further 27 random stations were later allocated in strata where catch rates were highly variable. This was done as the opportunity arose and not as a distinct second phase to the survey (Francis 1989). The position of each sampling station within each stratum was selected randomly by the use of a custom computer program. All positioning in this survey was carried out with a Magellan PRO 1000 Global Position System receiver (GPS). Dredge Sampling The 6-m vessel Susie, used for the population survey, was powered by twin three-stage Hamilton water-jet units. During dredging operations, one jet unit was used to propel the vessel while the other was used to provide water to the hydraulic dredge. A Japanese design hydraulic "Rabbit Dredge" was used in the sampling. The dredge was 1.3 m wide and constructed according to the design specifications in Michael et al. ( 1990, their Fig. 7). The hydraulic dredge operates by injecting seawater into the sand at a rate sufficient to liquefy the sand in front of the dredge and permit its easy towed progression through the substrate at a depth of approximately 200 mm. Clams >10 mm' in length were retained by fitting a 10-mm steel mesh screen across the top frame, side grills, and filtration grill area posterior to the digging bit, and the catch bag was fitted with a 10-mm knot-to-knot liner so that small clams would be retained once caught. To ensure comparability of samples col- lected on different occasions, all dredging was restricted to calm seas with swells of less than 0.75 m. The latitude and longitude of the start of each standard 50-m tow were recorded with the GPS receiver. Depth, to the nearest 0.2 m, was determined by echo sounder. Depths at the start of each tow were converted to chart datum by the use of the appro- priate published tide table corrections. Where a randomly selected station was due to start within 50 m of a stratum boundary, and extend in a direction where the tow might cross the boundary should the start position be in error, the tow direction was con- strained to be into the stratum. This only occurred in 4 of 127 tows, so a systematic error is unlikely. By detailed position plot- ting, it was found that none of the 127 samples used in the analysis crossed a stratum boundary. To measure distance towed, a lead line attached to a 3-kg weight, with weights approximately every 200 mm, was thrown overboard to mark the start of the tow once operational water pressure was reached. During the tow. the lead line was continu- ously paid out by hand and the tow was stopped after precisely 50 m of the weighted line had been paid out in a straight line over the bottom (allowance was made for the water depth). The area swept was thus constant at 65 m" (1.3 x 50 m). Propulsion was adjusted and controlled to achieve slow but approximately constant speed over a period of not less than 5.0 min for the 50-m tow. Tow duration was recorded on each occa- sion. Care was taken to ensure that all tows were made in a straight line. Tow orientation varied primarily according to the direction of along-shore current flow and sea breeze. For safety reasons, tow orientation generally headed into any swell and so tended to be offshore. Every effort was made to standardize the manner of dredge operation with the aim of obtaining consistent dredge per- formance. No tows were made where the dredge was retrieved completely full, although in the sand dollar-rich areas longer tows would have been full. For each tow, the surf clams were separated from other mac- rofauna and placed in labelled bags; all samples were stored in large insulated bins containing frozen coolant pads until the end of the sampling day. Sand dollars were counted aboard before being returned to the sea. Sample Processing After landing, all samples were frozen until laboratory process- ing, which occurred usually within 1 wk of the date of sampling. The surf clams from each tow were separated into the five species, counted while defrosting, and weighed by species to the nearest gram while still partly frozen so as to avoid water loss. The max- imum anterior-posterior length of each surf clam was measured to the nearest 1.0 mm with vernier callipers. Data Analysis Standard methods were used when estimating biomass and population sizes from the stratified survey design (Snedecor and Cochran 1967). In these analyses, the data from each stratum were weighted according to relative stratum area. Length-frequency data from each stratum were also weighted according to relative stratum area and were used to describe the size structure of the five species of surf clam. This latter information suggests the extent of potential recruitment and the size ranges potentially available to the fishery. The surf clam species are thought to differ in their depth of burial in the sand and in their distribution with respect to river outlets and different sediment types. Thus, each surf clam species may differ in their vulnerability to being taken by the hydraulic dredge. This may also vary with water depth and sediment type. In practice, vulnerability to capture would be measured by estimating the dredge efficiency. Because dredge efficiency was difficult to estimate with sufficient precision to be useful at all depths and in all sediment types, a vulnerability (dredge efficiency) of 100% was assumed in the biomass and population estimates. This leads to conservative estimates. Density and Size Relationships With Water Depth The mean frequency per tow of each species was estimated for all stations in each of eight 1-m depth intervals: 1-2 m, 2-3 m, 8-9 m. Mean counts were plotted against depth in- terval to provide a visual indication of whether a species exhibited a modal depth band and. if so. where it lay. Similarly, the mean shell length of each species for each tow was plotted against depth of tow to indicate any relationship with depth. Linear regression lines were fitted to these data to indicate any trend present. These regression lines are only intended to suggest the depth trends because they are weakened by the pres- ence of larger animals in all depths where the species are found. Foi comparison, moving averages of average shell length against depth were also plotted. 334 Haddon et al. Distribution Patterns of Each Species The biomass of each of the five species in each of the tows was plotted against the geographical position of the tow to compare the relative distribution of each species. This was also done for the counts of F. zelandiae. To assess the degree of aggregation visu- ally, frequency versus abundance histograms were plotted and inspected. These were compared with the total weighted variance to mean ratios of abundance. Finally, to produce a broad overview of the degree of mixing of species, the cooccurrence of species per tow was determined by collating the number of species per station for all stations. TABLE 2. Population Numbers for Each of the Five Species of Surf Clam. RESULTS Biomass Estimate D. anus was the most abundant species by weight, and 5. aequilatera was the least abundant (Table I ). The total biomass of surf clams between the Manawatu River and the Waitohu Stream was estimated at 1032.698 t. This implies an average biomass of approximately 37.55 t km"' of coastline (assuming 100% dredge efficiency). D. anus contributed 47.6% of the biomass. followed by M. discors (18. 9%). which had only about a third of the D. anus abundance. P. donacina (16.5%) and M. murchisoni {\A.\%) contributed very similar proportions of the overall catch, followed by the least common species. S. aequilatera (2.9%-; Ta- ble 1). The relative proportions of individuals of each species differed from those of the biomass estimates. D. anus dominated even more strongly, with a population of approximately 28 million, whereas the remaining four species were within 1% of each other, with approximately 4 million individuals each (Table 2). M. dis- cors was slightly more common than the rest, but 5. aequilatera greatly increased its relative importance, indicating that its popu- lation was made up of mostly smaller, lighter individuals (Fig. 2). The coefficients of variation for the estimates of both biomass and population numbers were less than 20% for all five surf clam species. Length Frequency Distributions As indicated by their length frequency distributions, juveniles of all species were abundant except for P. donacina (Fig. 2). The only clear mode for D. anus was between the lengths of 43 and 60 mm. with a peak at 50 mm (Fig. 2). There are suggestions of numerous other modes at smaller sizes but none were clear. Three modal classes were found in M. discors (Fig. 2). only one of which was clearly marked. The first ranged between 7 and 24 mm, TABLE I. Biomass Estimates for Each of the Five Species of Surf Clam. .Average Standard Biomass Species gm - Error cv % Tonnes D. anus 12.76 1.7234 13.51 491.381 P. donacina 4.44 0.7982 17.96 171.105 M. murchisoni 3.77 0.4042 10.72 145.238 M. discors 5.07 0.7853 15.48 195.319 S. aequilatera 0.77 0.1508 19.58 29.655 Total 26.82 1032.698 Percent- Mean Standard Population age of Species Clams m~" Error CV % Size Total D. anus 0.7280 0.0799 10.97 28.034,578 63.46 P. donacina 0.0932 0.0163 17.47 3.587,200 8.12 M. murchisoni 0.1039 0.0086 8.30 4,002,290 9.06 M. discors 0.1220 0.0189 15.53 4,696,094 10.63 S. aequilatera 0.1002 0.0132 13.20 3,856,551 8.73 Total 1.1473 44.176.713 Data are from 127 stations, in 20 strata, CV, coefficient of vanation. centred approximately at 15 mm. the second ranged between 25 and 38 mm. centred at approximately 33 mm. and the third and largest ranged between 39 and 72 mm, centred at approximately 51 mm. However, these major classes appeared to be made up of a number of components. In M. murchisoni. there appeared to be three or four distinct size classes (Fig. 2), all of which had approximately the same abundance as the smaller modes of M. discors. The first was between 10 and 28 mm, centered on approximately 18 mm, the second was very broad and ranged from 28 to 58 mm. centered on 35 45 55 65 75 85 S. aequilatera n=818 15 25 35 45 M, murchisoni 55 65 n=877 75 85 Data are from 127 stations, in 20 strata. CV. coefficient of variation. 25 35 45 55 65 75 85 Shell Length (mm) Figure 2. Weighted shell length frequencies of each species of surf clam. The distributions are shown as a connected whole to emphasize which size ranges were most common. Note that the frequency scales differ between species. BiOMASS AND Distribution of Exposed Coast Surf Clams 335 approximately 46 mm (although this may be made up of two main component classes each with different modes), and the last class ranged from 59 mm up to 83 mm in length, centered on approx- imately 67 mm. For 5. aequilatera. the length frequency distribution was very noisy despite measuring 818 individuals (Fig. 2). Three possible major classes may be identified, with modal groups approximately in the centre of the range of each: the first between 7 and 13 mm, the second between 14 and 23 mm, the third between 24 and 38 mm. Beyond that, shellfish were found with lengths up to 54 mm but only in low numbers. With P. donacina, a small peak of new recruits centred on 15 mm was found between 7 and 25 mm (Fig. 2). Beyond that, there was only the final collective modal group, ranging from 55 to 80 mm, with a modal value of approximately 70 mm. Deplh-to-Size Relationships When average shell lengths for each tow are plotted against tow depth, the five species exhibit different distributions (Fig. 3). P. donacina is concentrated in waters <4 m in depth, but the mean size (length) of animals had no relation with depth so that the regression line was effectively horizontal. S. aequilatera had more smaller individuals in shallow waters than in greater depths, so the regression line had a slightly positive gradient (Fig. 3). M. discors exhibited a distribution that was the inverse of 5. aequilatera in that there were more smaller animals in deeper water so the re- gression line had a negative gradient. Finally, both M. murchisoni P. donacina 80 .;.T'.:-. •■ ^— 60 40 20 ■ P. donacina 0 50 - 40 30 20 ■ . . ." • . t * . ^.^^ -^^ •• . .• 10 S. aequilatera 0 80 ■ 60 ■ — '^^^•^^'W— -O?;^ • . •. • 40 ■\ .•.'. "^^^~^~^ 20 M.discors . •: • 0 80 ■ 60 I ' 1 * • ' '. ,*. . ** • . 40 t ^^^^7"=--^-^— -^— 20 M. murchisoni . • 0 50 - .... 40 - ' |i' ■ . , M.,J_^ .a_^_vj^.' . • 30 20 " * . • * • * • 10 D. anus "l 2 3 4 5 6 7 8 9 Depth (m) Figure 3. Average shell length of each tow plotted against chart datum depth for each tow, for each species. The lines are the least squares regression lines. 12 3 4 5 6 7 8 9 Depth (m) Figure 4. Average catch in numbers of shellfish per tow within 1-m depth categories showing the depth ranges within which the highest densities were found for each species. and D. anus had a pattern similar to that of M. discors. except that the negative regression lines were steeper because there were fewer larger animals in deeper water as well as fewer smaller animals in shallow water (Fig. 3). Thus, P. donacina and 5. ae- quilatera are both relatively shallow water species and the other three surf clam species are all relatively deep water species. The same trends were found when all individual lengths from each tow (instead of mean lengths) were plotted against depth; the diagrams, however, were less clear because of the profusion of points. Again, similar trends were found in the plots of moving average of average length versus depth, with the changes seen remaining gradual. Depth and Density Relationships Each species exhibited a different optimal range of depth, as defined by modes of numerical abundance (Fig. 4). These pre- sumably optimal depth ranges permit an ordering of species with increasing depth starting with P. doiuicina, having peak densities between 2 and 3 m and no animals deeper than 4 m. S. aequilatera was also found at its highest densities between 2 and 3 m but was also found out to 8 and 9 m, although in lower numbers. M. murchisoni had a broad mode with high abundance between 3 and 6 m in depth and slightly higher numbers at 3^ m. M. discors. on the other hand, exhibited a clear mode between 3 and 5 m in depth, although it was also found at relatively constant medium densities to depths of 9 m. Finally. D. anus had a clear mode lying between 336 Haddon et al. 5 and 7 m below chart datum (Fig. 4). although the species was also found in all depths. Changes in the average relative abun- dance of the different species were most obvious between 3 and 4 m. In shallower water, P. donacina and 5. aequilatera dominated catches, whereas below 3 m. the two Mactra species and D. anus were the most abundant, particularly the latter (Fig. 4). Geographical Distribution The distribution of biomass of the different species in relation to depth and geographical distribution was illustrated by plotting catch weight categories against position of tow (Figs. 5-7). Sta- tions with relatively high catch levels of D. anus were found along the whole of the surveyed coast, especially in the 5- to 7-m strata, and were absent from only three stations. Low- and high-density tows were often close together, indicating a high degree of aggre- gation or patchiness (Fig. 5). The two Mactra species were both generally distributed, with relatively low catch levels throughout the survey area, although there were two areas, at opposite ends of the survey area, where high catches of M. discors were made (Fig. 6). The average low catch levels of S. aequilatera combined with its wide distribution over the range of depths and latitudes included in this survey contrast with the concentration of relatively high catch levels of P. donacina in the inshore areas (Fig. 7). Although P. donacina was found along the full length of the surveyed coast, catches were highest in the south (Fig. 7a). The high catch levels of P. donacina along with none being caught in tows from water deeper than 4 m imply that this species also had a relatively high degree of aggre- gation. None of the areas of high or low catch levels were obvi- ously related to river or stream outlets, and a crude latitudinal gradient in biomass was apparent only in P . donacina. In terms of absolute densities over all stations of all surf clams combined, the 43.5 1 - D anus 1 1 1 g m' -) .•D./ O >.^0 - O <30 -0" / 'o » / 0 <:o .9-1 • < 10 sCh = 0 • o °/ \- / : ^ s r - (§/ - - -QJ - - o 9-/" - .o° 7 .S)y J . o ° ./ \ 1 1 1 6.5 7 5 8,5 9 5 10 5 115 12 5 65 7 5 8.5 9.5 10 5 115 12 5 lx)ngitude Minutes Longitude Minutes Figure 5. Schematic diagram of the surveyed coast. Axis values relate to latitude 40°S and longitude 175°E, respectively. The strata outlines indicate the 20 strata used in the survey. The inshore strata are be- tween 1 and 3 m, the next are .^ and 5 m, then 5 and 7 m, and finally 7 and 9 m. i). anus illustrates the distribution of different densities by weight. Units are grams per square meter. 43 5 6.5 7.5 8.5 9.5 10.5 11.5 12.5 6.5 7.5 8.5 9.5 10,5 115 12,5 Lx?ngitude Minutes Longitude Minutes Figure 6. Schematic diagram of the surveyed coast. Axis values relate to latitude 40°S and longitude I75°E, respectively. The distributions of different densities by weight of both A/, discors and M. murchisoni are indicated. Units arc grams per square meter. average density was 1.232 animals m ", and the minimum den- sity found at a station was 0.015 clams m~-, whereas the maxi- mum was 9,969 surf clams m"". A more obvious gradient of density of individual was observed with F. zelandiae (Fig. 8). These were distinctly more abundant in the southern parts of the survey area. s aequilatera o g m" >.30 o <.30 0 <20 • <10 = 0 43,5 65 7 5 8,5 9,5 10,5 11,5 12,5 6,5 75 8.5 9.5 10,5 115 12 5 Longitude Minutes Longitude Minutes Figure 7. Schematic diagram of the surveyed coast. Axis values relate to latitude 40°S and longitude 175°E, respectively. The distributions of different densities by weight of both P. donacina and S. aequilatera are indicated. Units are grams per square meter. BioMASs AND Distribution of Exposed Coast Surf Clams 337 43.5 6.5 12.5 7.5 8.5 9.5 10.5 11.5 Longitude Minutes Figure 8. Schematic diagram of the surveyed coast. The axis values all relate to latitude 40°S and longitude 175°E, respectively. The distri- butions of different densities of counts of F . zelandiae are indicated. Units are individuals per square meter. Other Indications of Spatial Distribution All five species of surf clam exhibited highly positively skewed frequency versus abundance histograms, suggesting they were ail highly aggregated. A comparison of the ratio of the overall weighted variance to mean densities (Table 3) indicated that the most highly aggregated species was D. amis, with P. donacina next, followed by M . discors. then M. miirchisoni . and finally S. aequilalera . This is consistent with the visual assessment of the distribution maps (Figs. 5-7). Cooccurrence of Species Eight of 127 stations contained only a single species of surf clam. One of these yielded only P. donacina (a single individual), TABLE 3. The Total Weighted Variance, Mean, and Variance/Mean Ratio for Biomass (g m ^) for Each Species. Species Mean Variance Variance/Mean D uiui.s 12.76 12.549.3 983.5 P donacina 4.44 2.691.8 606.3 M. discors 5.07 2,605.6 513.9 M. murchisoni iJl 690.2 183.1 S. aequilalera 0.77 96.1 124.8 and the other seven contained only D. anus. None of these eight stations had high densities of surf clams. Eighteen tows contained two species, which was similar to the 17 that contained five spe- cies, whereas the most common situation was where a tow con- tained either three or four species (Fig. 9). DISCUSSION Biomass Estimation This survey was the first large-scale (27.5 km of coast), strat- ified random survey of suif clams on the Wellington West Coast. A total biomass of 1,032 t implies that the average biomass per linear kilometre of coast (between 1 and 9 m chart datum) is approximately 37.5 t km" '. This value is lower than the average of the three estimates made by Cranfield et al. (1994), but it is greater than their lowest estimate. Our estimate is likely to be conservative because of the assumption of 100% dredge effi- ciency. Michael et al. (1990) calculated a dredge efficiency of 65% with respect to a predominantly P. donacina bed containing approximately 58 surf clams m"". This was up to 10 times the maximum densities found elsewhere (Cranfield and Michael 1987) and over 5 times more dense than the maximum and 47 times more dense than the average density of surf clams found in this study. The relationship between the rabbit dredge's catching efficiency and surf clam density is unknown. Other hydraulic dredge designs have had catch efficiencies estimated between 80 and 100% (Meyer et al. 1981. Smolowitz and Nulk 1982). Notwithstanding adjustments with respect to dredge efficiency, the total biomass along the surveyed stretch of coast is lower than the earlier sam- pling suggested might be present. The possible harvest from this resource will be correspondingly limited, depending on the rela- tive productivity of the species concerned. Potential Recruitment A number of size classes, which with some relatively fast growing species, especially S. aequilalera and M. discors, may relate to age classes (Cranfield et al. 1993). were found in all species except P. donacina. Recruitment of juvenile size classes is clearly occurring in the deeper water species. Very few juveniles of P. donacina were found, indicating that these occur elsewhere. Adults of the related P. sublnangutata occur mostly between the low-tide line and 1.0-m chart datum. Juveniles of that species OV) ' 50- - 36 41 17 o 40 - - |30- £ 20- - 18 10 - 8 0 - The larger the ratio the greater the degree of aggregation found in the samples. 12 3 4 5 Species per Tow Figure 9. Relative frequencies of tows containing different numbers of species of surf clam. The numbers contained above each bar are the number of tows in each category. 338 Haddon et al. mainly occur towards the high-tide mark and migrate to lower levels only when they have attained a size of approximately 25 mm (unpublished data). Juvenile P. donacina have also been observed both intertidally and in shallow subtidal water. It appears that prerecruits of P. donacina mainly settle and grow in water shal- lower than 1.0-m chart datum. Their absence from the observa- tions of length frequency are therefore to be expected. However, why no P. donacina were observed between approximately 20 and 40 mm is unknown. Distribution of Shellfish Lengths Relative to Depth The manner in which the mean length of shellfish varied with depth reflected whether the species belonged to either the inshore group (P. donacina or S. aeqidlatera) or the offshore group (the two Mactra species and D. anus). The lack of any relation be- tween the depth and mean length of P. donacina reflects both its restricted depth distribution and the fact that its juveniles develop inshore of the 1.0-m survey boundary. It implies that as they grow in length, juveniles of this species must, at some stage. migrate offshore to the adult beds. S. aequilatera has a far wider depth distribution, but smaller individuals were found predomi- nantly in shallower water. This species is also found in water less than 1 .0-m chart datum. Clearly, some of the smaller animals must be moved by wave action or migrate offshore as they develop and grow. The reverse is true for the two Mactra species and for D. anus. Because smaller individuals of these three species are more common offshore, then to maintain the distribution of larger ani- mals across all depths, some of the smaller individuals must move inshore as they increase in size. In both M. munhisoni and D. anus, there is some indication that larger animals are also less common in the deepest stations, which implies such movement. With S. aequilatera and M. discors. on the other hand, the move- ment of individuals would only be facultative. An alternative hypothesis, that the observed distributions were brought about by differential settlement or the survival of larvae over a number of years, is less likely. If this were the case, all species might be expected to exhibit the same distribution of size versus depth. Depth and Density Relationships There is a succession of modes of maximal relative abundance from shallow to deep for the five different species. D. amis dom- inated the absolute numbers in all depths except the shallow range of 1-3 m, where P. donacina predominated. Despite the clear succession in maximum relative abundance, there was a great deal of overlap between all species (except P. donacina. which is re- stricted to shallow water) across all depths, especially out to 7 m. In waters deeper than 7 m, the numbers of surf clams were rela- tively low, but at least medium densities were found at all depths between 4 and 7 m. The most marked division of surf clams is the relatively shallow water (1- to 4-m chart datum) group of two species (S. aequilatera and P. donacina) and a deeper water (4- to 8-m chart datum) group of three (A/, murcliisoni. M. discors. and D. anus). In depths where potentially commercially viable (fish- abie) densities occur, it appears that it would be almost impossible to fish without catching a mixed bag of surf clams. However, without further information concerning how the different species are distributed along a more extensive section of the coast, no general conclusions can be made as to whether it is possible to target a particular species at all sites along the coast. Geographical Distribution by Density of Surf Clams The domination of the biomass by D. anus and the concentra- tion of this species in the 3- to 7-m depth strata is clear from the distribution maps (Fig. 5). The close proximity of stations with high levels of biomass to those with very little biomass indicates the high degree of patchiness in this species. The two Mactra species are more evenly distributed and are well mixed with the stations containing D. anus. Visually, these three species appear to form a well-mixed group that predominates in the 3- to 7-m strata. The inshore group differs from the offshore group in that its members are very different in their abundance and relative distri- butions. P. donacina is concentrated in the shallow strata and has a gradient of increasing density from north to south. S. aequilatera differs in being at consistently low densities throughout the survey area, with increasing numbers of zero count stations in the deeper strata. The distribution bubble maps indicate the mixed nature of the communities present at any single location along the coast. The geographical information is consistent with the information con- cerning depth distribution in that both imply that target fishing would be difficult. It might be possible to locate a patch of surf clams rich in a particular species, especially with the inshore group, but it would be very difficult to fish extensively without obtaining high proportions of at least two other species. Thus, the closest approach to target fishing would be to concentrate effort within patches known to have high densities of the species of interest. Generally, however, the distribution maps and high levels of aggregation suggest that one would only be able to dredge a short distance before leaving a concentrated patch of a particular species. The Effect of Fellaster zelandiae on Surf Clam Fishing In the northern parts of the survey area, the presence of the New Zealand sand dollar was only a minor nuisance because it was present in relatively low densities. In the southern area, however, their high densities (sometimes more than 2.000 individuals in a 65-m" tow) meant that they became a severe nuisance, hampering fishing operations by clogging the catch net with this nontarget species. Commercial fishing operations would either have to find an efficient means of separating the surf clam catch from usable surf clams or avoid areas of high densities of sand dollars. The extent and distribution of this species should therefore also be examined when developing a New Zealand surf clam fishery. Management Options Tagging experiments have demonstrated that growth rates vary considerably between surf clam species (Cranfield et al. 1993). Estimates of size-at-age have been made that indicate that, for the Wellington West Coast, 5. aeciuilatera (4 y) and M. murcliisoni (5 y) grow the most quickly to reach a "marketable size,"" whereas D. anus ( 14 y) grows the most slowly (Cranfield et al. 1993). The growth rate for the first two species suggests relatively high levels of natural mortality, which, in turn, implies that these species could sustain a relatively high fishing mortality with minimal risk of harming the stock. Unfortunately, it would be necessary to be able to target for particular species when fishing before being able to manage the surt' clam species separately. With the low catch rates possible along this coast, it is likely that continuous towing with some mechanism bringing the catch to the surface for continuous sorting would be necessary for the fish- BioMASS AND Distribution of Exposed Coast Surf Clams 339 ery to be economical. This is a further reason why it would be both difficult and impractical to target fish for particular species. To date, only the East Coast United States surf clam fishery uses a total allowable catch (as quota) to control the fishery. The U.S. fishery is for Spisulti soliclissinui. which is relatively long lived and slow growing and has variable recruitment. Accord- ingly, the annual quota is set at a low proportion (0.045) of the recruited biomass (Murawski and Serchuk 1989, Cranficid el al. 1994). Setting a total allowable catch appears to work with the United States Atlantic Coast surf clam fishery. However, the patchy na- ture of the distribution of surf clam species along the Wellington West Coast, combined with the reported differences in productiv- ity and growth, as well as the occasional beach strandings of large numbers of surf clams, implies that the New Zealand surt^ clam community may be more dynamic in yield and composition than the U.S. East Coast situation. If the relative proportions of the different surf clam species can alter markedly from year to year, then managing the fishery by the use of a constant catch strategy, such as a maximum constant yield, to set total allowable catch levels for each species may not be the best strategy. It is suggested that because New Zealand surf clam populations appear to be dynamic in terms of both population size and relative proportion of species, a management strategy closer to a constant fishing mor- tality be adopted. To implement this, the fishable biomass of each species within a fishing area would need to be determined, before each annual fishing season, and each year's survey would need to be used to determine the potential catch. ACKNOWLEDGMENTS Funding for this work came from the Maori Business Technol- ogy Initiative, Foundation for Research, Science, and Technol- ogy. We gratefully acknowledge the efforts of the Shore Crew assistants; Mr. Jacko and Mrs. Jo Eru-Te Hopu, Ms. Brenda Pratt, Mrs. Kim Ellison, Mr. Brett Wilson, and Mr. Delson Packer. We also thank the crew of the R.V. Susie, especially Mr. Steven El- lison (Captain) and Mr. Bruce McKelvey; Mr. William Ellison, Mr. Terrence Waka, Mr. Wayne Eru-Te Hopu, and Mr. Kina Wylie for all of their efforts under often difficult circumstances. Finally, we are especially grateful to Mr. Russell Packer, Mr, Alan Morgan, and Mr. Willie Packer for their cooperation and kindness in our times together. LITERATURE CITED Anderlini, V. C. & R. G. Wear. 1991. Surf clam resource survey — Manawatu to Rangitikei Rivers. Via. Univ. Coast. Mar. Res Unit Rep. 16. 35 pp. Butts, D. 1982a. Faunal identifications from Muhunoa west midden (N152/50), Horowhenua. N.Z. Archaeol. Assoc. Newsl. 25:191-194. Butts, D. 1982b. Preliminary observations relating to the coastal middens of Manawatu. N.Z. Archaeol. Assoc. Neusl. 25:268-276. Caddy, J, F. (ed.). 1989. Marine Invertebrate Fisheries: Their Assessment and Management. J. Wiley, New York. 624 pp. Carkeek, W. C. 1966. The Kapiti Coast. Reed, Wellington. 187 pp. Cranfield, H. J. & K. P. Michael. 1987. Surf Clam Resource, Cloudy Bay, Marlborough. N.Z. Fish. Intern. Rep. No 75. II pp. Cranfield, H. J. & K. P. Michael. 1992. Surf clam,s — a fishery in waiting? N.Z. Prof. Fish. 6:57-60. Cranfield, H. J. & K. P. Michael. 1993. Surf clams— a fishery still in waiting. N.Z. Prof. Fish. 7:41^3. Cranfield, H. J., Michael, K. P. & D Stotter, 1993. Estimates of growth, mortality, and yield per recruit for New Zealand surf clams. N.Z. Fish. Assess. Res. Doc 93120. 47 pp. Cranfield, H. J., Michael, K. P., Stotter, D. & I, J. Doonan. 1994 Dis- tribution, biomass and yield estimates of surf clams off New Zealand beaches. N.Z. Fish. Assess. Res. Doc. 9411. 27 pp. Francis, R. I. C. C. 1989. A standard approach to biomass estimation from bottom trawl surveys. N .Z. Fish. Assess. Res. Doc. 8913. 4 p. Meyer, T. L.. Cooper, R. A. & K. J. Pecce. 1981. The performance and environmental effects of a hydraulic clam dredge. Mar. Fish. Rev. 43:14-22. Michael, K. P.,G. P. Olsen, B. T. Hvid&H. J. Cranfield. 1990. Design and performance of two hydraulic subtidal clam dredges in New Zealand. N.Z. Fish. Tech. Rep. No 21. 16 pp. Morton. J. E. & M. C. Miller. 1973. The New Zealand Sea Shore. Col- lins, Auckland. 653 pp. Murawski, S. A. & F. M. Serchuk. 1989. Mechanized shellfish harvest- ing and its management: the offshore clam fishery of the Eastern United States, pp. 479-506. In: J. F. Caddy (ed.). Manne Invertebrate Fisheries: Their Assessment and Management. J. Wiley, New York. Powell, A. W. B. 1979. New Zealand Mollusca. Collins, Auckland. 500 pp. Smolowitz, R. J. & V. E. Nulk. 1982. The design of an electrohydraulic dredge for clam surveys. Mar. Fish. Rev. 44:1-18. Snedecor, G. W. & W. G Cochran. 1967. Statistical Methods. 6th ed. The Iowa State University Press, Ames. 593 pp. Journal of Shellfish Research. Vol. 15. No. 2. 341-344. 1996. EFFECT OF GROWOUT DENSITY ON HERITABILITY OF GROWTH RATE IN THE NORTHERN QUAHOG, MERCENARIA MERCENARIA (LINNAEUS, 1758) JOHN W. CRENSHAW, JR., PETER B. HEFFERNAN,' AND RANDAL L. WALKER Shellfish Aquaculture Laboratory Marine Extension Service Universit}' of Georgia 20 Ocean Science Circle Savannah. Georgia 3141 1- 101 1 ABSTRACT Realized hentability tor the increase in the rate of growth in the northern quahog. Mercenaria mercenaria. was determined under conditions of moderate growout density (<90 per sq. ft.) independently for two lines by Crenshaw et al. (1993). For one line, termed Group A, a mean estimate of heritability of 0.402 was obtained. This estimate was based on a single generation of selection in which a standardized selection differential (i) of about 1.5 standard deviations was used, representing a selection intensity of about 16%. When Select and Control progeny of Group A were maintained in growout at high densities (>350 per sq. ft.). Control progeny grew at significantly greater rates than Selects, thus resulting in negative estimates of realized heritability. Clam slocks were collected in House Creek. Little Tybee Island. Wassaw Sound, in coastal Georgia. Same-age cohorts of F, progeny were established in Apnl and May 1986. Progeny were reared in the laboratory until December 1986. when they were transferred to growout cages in an intertidal creek. Selection was earned out for F, cohorts in March 1988. Select and Control parental groups were identical in number, the latter randomly chosen from the entire population before ascertaining the Select-line parents. Spawnings of the Control and Select Group A F, parents occun-ed on July 6 and 7, 1989. respectively. Clam progeny were transferred from nursery to growout cages on September 12, 1990, and density was reduced for Group A moderate-density cohorts on April 1, 1991. Other Group A cohorts of Select and Control lines were maintained at high density in growout cages similar to those holding moderate-density cohorts. For clams reared at moderate density, heritability estimates were calculated on September 12. 1991. for clams of Group A. Group A clams maintained in high-density growout developed more slowly than those reared at moderate density, and it was not appropnate to take measurements for estimating heritability until March 16. 1992. KEY WORDS: Genetic selection, hentability. growth rate, crowding, quahog INTRODUCTION Quantitative genetic selection may be used in a living organism to improve any trait for which there exists additive genetic vari- ance. The process is particularly useful in that selection may be carried out over many generations, with progress resulting in each, until genetic variance for the trait is e.xhausted. Newkirk ( 1980) reviewed genetic research involving commercially important bi- valves, emphasizing the potential of selective breeding. Humphrey and Crenshaw (1989) reviewed subsequent genetic research in- volving bivalve shellfish, emphasizing the genetics of the northern quahog. Mercenaria mercenaria. and outlined simplified proce- dures that could be used to develop estimates of realized herita- bility with shellfish. The realized heritability estimate resulting from genetic selec- tion is particularly useful as "an empirical description of the ef- fectiveness of selection" (Falconer 1981). This estimate can be used to estimate the number of generations of selection of a given intensity that would be required to achieve a particular goal. In a mariculture operation, successful selection for rate of growth in shellfish would result in reduction of time in growout. reduced rearing expenses, and improved profitability. Efforts to estimate a realized heritability for growth rate in the northern quahog or hard clam were initiated in our laboratory a number of years ago. Sub- sequently, estimates of heritability of growth rate in the hard clam have been published by ourselves and others (Hadley 1988. Had- ley et al. 1991. Crenshaw et al. 1993). 'Present address: Marine Institute. 80 Harcourt Street. Dublin 2. Ireland. Many factors may affect the outcome of a given selection pro- gram. We have reported on negative growth and survival effects in the larval and embryonic progeny of northern quahogs and of bay scallops selected for rapid growth rate (Heffeman et al. 1991, Heffeman et al. 1992). However, reduced larval rate of growth in both cases was more than offset by subsequent high growth rates in juveniles and young adults. The effect of environment on selection response is unpredict- able. It is well known that selection carried out in a given envi- ronment may produce different results in different environments (Falconer 1981). The heritability of a trait estimated in a given environment predicts response to selection most accurately in that same environment. In this study, we demonstrate the importance of crowding or population density on response to selection. MATERIALS AND METHODS Two lines of M. mercenaria. here tenned Groups A and B, were established form mass spawnings of wild stock clams from House Creek. Little Tybee Island. Wassaw Sound, in coastal Georgia, on April 4 (Group A), and May 8 (Group B). 1986. This report deals only with cohorts of Group A. Postsettlement (juve- nile) stages were reared in downwelling nursery systems following standard nursery culture procedures for bivalve molluscs (Casla- gna and Kraueter 1981. Heffeman et al. 1988). In order to retain all genetic variance possible for rate of growth, cohorts were never subjected to size culling, as is standard practice in commercial hatcheries. Larval and juvenile progeny were reared in the laboratory in ambient filtered seawater with food provided by Wells-Glancy 341 342 Crenshaw et al. cultured phytoplankton, supplemented by single-species algal cul- tures until December 1986. when they were transferred to growout cages at densities of approximately 1 .000 clams/m" in a sheltered, intertidal creek (House Creek) in Wassaw Sound, GA. Selection was carried out at approximately 2 y of age and mean shell length of about 32 mm for F, cohorts of Group A on March 16, 1988. At that time, densities had been reduced by mortality and escape to about 52 and 24 clams per sq.ft. for the two cages involved. Spawning of 177 Select parents and an equal number of Controls was induced by thermal stimulation. The Control parents had been randomly chosen from the entire population before the designation of the Select-line parental group (Crenshaw et al. 1988, Heffeman et al. 1991). Select-line parents were identified as those clams within the upper 16.6% of the shell length distribution, thus lying at or above a truncation cutoff point of 0.97 standard deviations above the mean. The mean shell length of Select-line parents was 1 .50 standard deviations above the overall mean of the population from which they were taken. Reared under the same conditions as the previous generation, F, Group A progeny were transferred from nursery to growout cages in the field on September 12, 1990. On April 1, 1991, density was reduced by dividing clams in each single cage approx- imately equally between two cages of the same size, resulting in a decrease in density from less than 90 clams per sq. ft. to less than 30 clams per sq. ft., termed here moderate density. On September 5. 1991, F2 clams of Group A Control line reared at moderate density were found to be approximately the size of the F, parental array at the time selection was carried out, and F^ Select- and Control-line measurements were made. It is on the basis of these Select- and Control-line groups, reared at moderate density, that our first estimates of realized heritability for growth rate in the northern quahog were computed (Crenshaw et al. 1993). On March 11, 1991. shortly before densities were reduced for Group A to bring about moderate-density conditions for the de- termination of heritability of growth rate, three growout cages were established for Group A Select and one for Group A Control populations to determine the effects on growth of very crowded conditions. The growout cages were wooden boxes, 4' x 4', covered with plastic-covered wire mesh. Established with 5,700 animals in each cage, density was initiated at slightly less than 360 per sq. ft. (approximately 3,200 clams/m~) for clams reared under these high-density conditions. On March 6, 1992, after 18 mo in growout, sample measurements of Controls grown at high density indicated that they had reached approximately the same size as clams grown at moderate density at the time they were measured for the purpose of determining heritability. Thus, clams grown at high density required about 6 mo longer to reach this size than was required for clams grown at moderate density. At that time, den- sities had been reduced by mortality and escape to between about 80 and 130 clams per sq. ft. On March 16, 1992, random samples of 200 from each cage were removed and measured to determine the effect of high density on heritability. RESULTS Mean shell length measurements and selection statistics of the F| generation of Group A are provided in Table 1 . As indicated above, a cutoff point of approximately 0.97 standard deviations above the mean was used, meaning that, after random selection of a Control group of F, parents, all remaining clams with shell lengths equal to or exceeding the mean plus 0.97 standard devia- tions were chosen to comprise the Select F, parental group. For the overall F, population, at 2 y of age, the weighted mean shell length of two replicates was 31.27 mm. The mean of the Select- line parents was 42.42 mm or 1.50 standard deviations above the mean of the population from which Select- and Control-line par- ents were taken. This figure of 1.50 represents the standardized intensity of selection (i) or "reach" for Group A. Mean shell length measurements of the F, generation of Group A, Select and Control lines, moderate- and high-density groups, are provided in Table 2. Response to selection or "gain" is estimated by the difference between the means of the F, progeny from F, Select and Control parents. For Group A, F, progeny of Controls reared at moderate density were found to have a mean shell length of 3 1 .97 mm; those of Select parents also reared at moderate density had a mean shell length of 35.57 mm. The difference, 3.60 mm, is highly signifi- cant by analysis of variance (ANOVA) (P < 0.0001) and is equal to 0.61 standard deviations of the Control progeny distribution (Table 2). This figure represents an estimate of response to selec- tion or "gain" for four growout replicates of Group A. Realized heritability (h") may be computed as the ratio of re- sponse ( = "gain") to selection ( = "reach"). For Group A clams reared at moderate density, it is estimated that h" = 0.613/1.499 = 0.409. For clams reared under crowded conditions in growout, the situation is quite different. F, progeny of Controls were found to have a mean shell length of 35.63 mm, whereas progeny of Select parents also reared at high-density conditions had a mean shell length of the only 33.78 mm. The difference, 1.85 mm, is highly significant by ANOVA (P < 0.0001), but clearly is in the wrong direction to represent a positive response to selection. The shell length of two of the three replicates of progeny of Select parents was significantly different from one another (Turkey's HSD Mul- tiple Comparison), but all were lower than the mean shell length of Controls, and we have lumped them for simplicity in our analysis. A calculation of heritability on the basis of the "gain" to "reach" ratio would produce a negative figure, which would be meaning- less. On March 6, 1992, 10 d before measurements were taken from TABLE I. Size attained by F, generation same-age populations of the northern quahog at time of selection in Group A Standard Mean shell Length Deviation Cutoff Point Mean of Group Number Measured in mm ± SE (SD) Above Mean Selected Parents Replicate 1 Replicate 2 1,232 842 390 31.27 ± 0.20 31.10 ± 0.24 33.05 ± 0.37 7.14 0.97 S.D. 42.42 mm (mean -I- 1.499 SD) QuAHOG Crowding and Heritability of Growth Rate 343 TABLE 2. Size attained in the northern quahng by F, generation progeny of F, generation Select and Control parents of Group A, reared under conditions of high and moderate density, at approximately the same size as their parents at the time selection for increase in growth rate was carried out. Standard Date of Number Mean Shell Length Deviation Difference in mm/ Gain in SD Density Group Measurement Measured in mm ± SE (SD) Signiflcance of Control Moderate A Control 9/09.91 200 31.97 ± 0.42 5.87 3.60/ 0.613 A Select 9/09/91 800 35.57 ± 0.23 6.60 P < 0.0001 Replicate 1 200 37.08 ± 0.44 Replicate 2 200 36.82 ± 0.45 Replicate 3 200 34.55 ± 0.46 Replicate 4 200 33.81 ± 0.47 High A Control 3/16/92 200 35.63 ± 0.42 6.0 -1.85/ -0.308 A Select 3/16/92 600 33.78 ± 0.27 6.55 P < 0.0001 Replicate 1 200 32.76 ± 0.47 Replicate 2 200 34,36 ± 0.44 Replicate 3 200 34.21 ± 0.47 the clams reared at high density, counts indicated that the mean sui-vival of three Select-line repHcates was 58.04%, as compared with 67.49% survival for the Control group. However, one of the Select-line replicates had a survival rate of 76.68%. DISCUSSION As reported earlier, the estimate of heritability of growth rate for Group A reared at moderate density, about 0.40. may be in- terpreted to indicate that about 40% of the total phenotypic vari- ance in growth rate is attributable to the average effects of genes under the environmental conditions of the experiment (Crenshaw et al. 1993). The selection event for the Group A F, generation, which was spawned to produce both moderate- and high-density F, groups, was carried out under conditions of moderate density. The density factor was introduced for the F, generation after about half a year in growout, and clams reared at moderate density remained in growout another 6 mo before measurements were made to calculate the heritability of growth rate. Clams of the high-density growout group, by contrast, remained in growout another year before reaching the same approximate size. The ef- fect of crowding is clearly indicated by the relative performance of Controls in the two situations. It took nearly twice as long for clams reared under high-density conditions to reach the same size as clams reared at moderate density. More interesting is the growth performance of Group A clams selected for rapid growth under two density conditions relative to their appropriate controls. The group reared under conditions of moderate density, approximating the conditions under which their parents were grown before selection, showed the expected re- sponse to selection by exhibiting more rapid growth than controls, but not quite as rapid as their parents, which had been selected as the most rapidly growing members of their cohort. The progeny of clams selected for rapid growth, when reared under conditions of high density in growout. by contrast, showed slower growth than Controls reared at high density indicating that something about the array of genes that determines rapid growth under conditions of moderate density actually has a negative effect on growth rate under conditions of high density. We suggest that selection for allelic combinations that foster rapid growth in mod- erate-density growout. which could be fairly termed nearly opti- mal conditions, will, at the same time, select against genotypes that perform best under conditions of stress. The control groups in the situation described here are very close genotypically to wild stocks, which have been fine tuned over thousands of generations of natural selection to withstand a variety of stressful conditions that occur in the natural clam habitat. It is reasonable, then, that a Control line, when subjected to stressful crowding, would do bet- ter at surviving and growing than a line selected for rapid growth, where there is an abundance of food and a low concentration of waste products. Although there was found fairly high survival in one Select-line replicate, the mean survival of Select-line repli- cates was less than that of the Control group. Falconer (1981. p. 322) prefers to regard the same trait as measured in two different environments not as one but as two different characters. His argument is that the "physiological mechanisms [involved] are to some extent different, and conse- quently the genes required for high performance are to some extent also different." In our original report of a determination of heritability of growth rate for the hard clam (Crenshaw et al. 1993). we discussed the related situation of a low estimate of heritability obtained from a second line. Group B. in which it could be shown that the low estimate was associated with the last 5 mo of growout. a period during which our data showed an appreciable reduction in the rate of growth of Select-line progeny relative to the growth rate of Control-line progeny. We attributed this difference in growth rates as due most likely to one or a combination of stressful factors associated with this period: 1 ) a change of tidal creek environment at the beginning of the period, or the new environment itself; 2) several stays in the laboratory with marginal food and low salinity at the beginning of the period; or 3) excessive population manip- ulation associated with the laboratory stays or the change of res- idence. The critical lesson for the mancultunst desiring to use stocks selected for desirable traits is that the selection process itself should be carried out under conditions comparable to those exist- ing in the hatchery, nursery, and growout facilities in which they will be used. It is well known that response to selection is maxi- mized in groups reared under conditions most similar to those under which their parents were reared and that stress of any sort tends to affect selected lines more profoundly than unselected lines (Falconer 1981). 344 Crenshaw et al. It has been pointed out that in this research we are not actually selecting for growth rate broadly in the quahog but. to be precise, for growth rate to a given size, as reflected by shell length. The result is, however, the same. Indirectly, we are selecting for growth rate, admittedly up to a specific size. In selection programs involving warm-blooded domestic animals, it is often fairly simple to select for a character at a specific stage of development, e.g.. "at birth" or "at one year of age." With invertebrate poikilo- therms, growth and development may vary enormously depending not only on temperature, but also on food supply as well as other factors. We would have preferred to use "attainment of sexual maturity or spawnability" as an appropriate milestone, and in fact, we have picked target shell lengths because animals of this size are usually sexually mature and spawnable. This correlation between size and sexual maturity is generally accepted. Determination of sexual maturity for each individual clam would require cytological examination and would be both excessively time consuming and somewhat traumatic. In essence, then, we use target shell length as a reflection of a given stage of development, and we have found that it is both practical and effective. An important problem emerges from the considerable variance in the time at which different members of a same-age cohort will reach a given shell length. In this work, we have found that when the mean shell length of our line selected for increased growth rate reaches the target size, most of the cohort are sexually mature. However, at the same time, the mean shell length of the control cohort will be smaller, and significantly more members will not be spawnable. The problem that this poses for the determination of heritability and our solution has been discussed elsewhere (Cren- shaw et al. 1993). ACKNOWLEDGMENTS This work was supported by Georgia Sea Grant Project No. NA84AA-D-00072. The technical assistance of Dr. F. O'Beim and Mr. D. Hurley is gratefully acknowledged, as is the secretarial assistance of Ms. D. Thompson. LITERATURE CITED Castagna, M. & J. N. Kraeuter. 1981. Manuel for growing the hard clam. Mercenaria. Special Report in Applied Marine Science and Ocean Engineering No. 249. Virginia Institute of Marine Science, Gloucester Point, VA. 110 pp. Crenshaw, J. W.. Jr.. P B. Heffeman & R. L. Walker. 1988. Growth of the hard clam in grow-out cages in coastal Georgia. J . Shellfish Res. 7(3);547-548. Crenshaw, J. W.Jr.P. B. Heffeman &R L Walker. 1993. Hentability of growth rate in the northern quahog Mercenaria mercenaria (Lin- naeus, 1758). pp. 10-15. In: M. S, van Patten (ed.l. Insh-American Technical Exchange on the Aquaculture of Abalone. Sea Urchins, Lobsters and Kelp. Galway, Ireland. Publication Number CT-56-93- 05, Connecticut Sea Grant College, Univ. of Conn., Groton Falconer, D. S. 1981. Introduction to Quantitative Genetics 2nd ed Longman Press, London, 438 pp Hadley. N. 1988. Improving growth rates of hard clams through genetic manipulation. World Aquaculture l9(3):65-66. Hadley. N. H..R. T Dillon & J J. Manzi. 1991. Realized hentability of growth rate in the hard clam Mercenaria mercenaria. Aquaculture 93:109-119. Heffeman. P. B . R L. Walker & J. W. Crenshaw, Jr. 1988. Growth of Georgia Mercenaria mercenaria (L.) juveniles in an experimental scale downweller system. G. J. Sci. 46(3):174-18l , Heffeman. P B., R. L. Walker & J. W. Crenshaw, Jr. 1991 Negative larval response to selection for increased growth rate in northern qua- hogs Mercenaria mercenaria (Linnaeus. 1758). J. Shellfish Res I0( 1): 199-202, Heffeman, P. B., R. L Walker & J. W. Crenshaw, Jr. 1992. Embryonic and larval response to selection for increased rate of growth in adult bay scallops, Argopecten irradians concentricus (Say, 1822). J. Shellfish Res. ll(l):21-25. Humphrey, C. M., & J. W. Crenshaw. Jr. 1989 Clam genetics. In: J. J. Manzi and M. Castagna (eds.). Clam Manculture in North Amenca. Elsevier, Amsterdam. Chap. 13. pp. 323-356. Newkirk. G. F. 1980. Review of the genetics and the potential for selec- tive breeding of commercial important bivalves. Aquaculture 19:209- Journal of Shellfish Research. Vol. 15. No. 2, 345-347. 1996. CYTOLOGIC SEXING OF MARINE MUSSELS {MYTILUS EDVLIS) SHELLEY A. BURTON,' * GERALD R. JOHNSON,' AND T. JEFFREY DAVIDSON^ Depariments of ^Pathology and Microbiology and ^Health Management Atlantic Veterinary College Universit}- of Prince Edward Island 550 University Ave. Charlottetown. PEL CIA 4P3. Canada ABSTRACT Four hundred eighty mature marine mussels {Mytilus edulis) were collected m late April from a mussel lease in an estuar>' m Pnnce Edward Island. Canada. Squash preparations made of reproductive gland (mantle) tissue of each mussel were stained with Wnght-Giemsa stain and were exammed by one investigator using light microscopy for sex determination. A strip of mantle tissue was also removed from each mussel, fixed in \Q9t buffered fomialm solution, and processed for histologic evaluation. The sex of each mussel was determined independently by evaluation of the histologic preparations by a second investigator. The evaluation of cytologic preparations of mantle tissue was highly accurate in determining mussel sex. in that all 480 mussels (261 females and 219 males) were correctly identified when compared with the traditional standard of histologic sexing. No hermaphrodites were observed. The advantages of the cytologic procedure over traditional histologic processing include ease of sample preparation and evaluation, short preparation time, low cost, and sparing of tissue for other studies. KEY WORDS: Cytologic, histologic, marine mussel. Myiilus edulis INTRODUCTION The classification of nianne mussels (Mytilus edulis) as male or female cannot be detemiined by morphology, size, or shell color. The sex of individual mussels or groups of mussels may have effects on physiologic parameters and the biochemical content of tissues. At present, the standard method available for determining the sex of mussels is to fix reproductive gland (mantle) tissues in fixative solution and then use histologic processing and light mi- croscopic examination for the presence of oocytes or spermatozoa. Histologic processing is expensive and permanently alters the tis- sue sample. A biochemical procedure for sex determination eval- uating color change after the exposure of mantle tissue to boiling thiobarbituric acid has also been reported (Jabbar and Davies 1987). This procedure destroys tissue and involves the use of a caustic chemical. The cytologic evaluation of tissues is an estab- lished diagnostic procedure used in veterinary and human medi- cine. In this procedure, cells obtained from squash preparations of small amounts of solid tissue or from needle aspiration of tissues are spread on glass slides, air dried, stained and examined by light microscopy. Compared with histologic processing and biochemi- cal procedures, this procedure is rapid and inexpensive and re- quires minimal tissue. The goal of this study was to determine the accuracy of classifying the sex of marine mussels (M. edulis) using squash preparations of mantle tissue compared with sexing using traditional histologic processing and evaluation. MATERIALS AND METHODS Mussel Selection and Processing Four hundred eighty mature (5- to 7-cm shell length) marine mussels {M. edulis) were collected in late April from a mussel lease in an estuary in Prince Edward Island, Canada. The mussels *Corresponding author. were removed from their shells, and each mussel had a small amount (approximately 1 mm") of tissue removed from the mantle with a scalpel. The tissue was gently squashed, smeared on a glass slide, and allowed to dry at room temperature. Dried smears were stained using a Wright-Giemsa stain pack (Fisher Diagnostics, Montreal, Quebec. Canada), a Romanowsky type of stain rou- tinely used in cytologic processing (Tyler et al. 1992). Each mus- sel also had a strip (approximately 0.5 x 2 cm) of mantle tissue removed and placed into \0'7c neutral buffered formalin solution. After fixation in the formalin solution ( 1 day or more), the mantle tissues were removed. They were dehydrated through an ascend- ing series of graded ethanols. cleared in xylene, embedded in paraffin wax (Paraplast Plus; Oxford Laboratories, Sherwood Medical. St. Louis, MO), sectioned at 5-(im thickness with a Reichert Histostat rotary microtome (Reichert Scientific Instru- ments. Warner-Lambert Technologies, Inc., Buffalo. NY), and placed on glass slides. Sections were stained with hematoxylin and eosin, as is routinely done in histologic processing (Luna 1968). Determination of Sex A classification of sex as male or female was made by evalu- ation of the Wright-Giemsa-stained squash preparations by a cy- topathologist (S.A.B.). A classification of sex as male or female was made independently by the evaluation of paraffin-embedded tissue sections by a histopathologist (G.R.J.). Statistical Analysis The number of mussels to be evaluated (480) was chosen by the use of a sampling formula (Martin et al. 1987), in which we assumed a possible 5'7t error rate and in which we wished to be 93-97% confident of results. RESULTS Cytologic evaluation of mantle tissue squash preparations cor- rectly classified all 480 mussels (261 females and 219 males) when 345 346 Burton et al. compared with independent histologic evaluation. The distinction between sexes was easily made by the evaluation of stained smears, because preparations from male mussels had numerous small, angular, dark purple spermatozoa seen (Figs. 1 and 2), and the preparations from female mussels had large, thin-walled, pale pink oocytes observed (Fig. 3 and 4). Interestingly, the squash preparation slides could even be roughly classified as male or female by visual inspection without microscopy, because stained smears from the male mussels had a blue tinctorial quality com- pared with stained smears made from the female mussels, which were purple. This was an incidental observation, and determina- tion of the accuracy of stain color to determine the se,\ of the 480 mussels was not performed. DISCUSSION Cytologic sexing of marine mussels {M. edidis) was 100% accurate in identifying the sex of individual mussels as compared with the traditional standard of histologic evaluation in this study. The cytologic evaluation was very simple, requiring only a few moments of scanning the smears to make the differentiation of male or female. No obvious mussels that could be classed as hermaphrodites were observed in our study. Hermaphrodites may occur at very low prevalence in mussel populations and may cause inconsistencies in cytologic versus histologic evaluation if present. The mussels in this study were evaluated in late April. At that time in eastern Canada, water temperatures are still cold and mus- sels are sexually mature but have not yet spawned. A similar study comparing cytologic and histologic sexing of marine mussels may be less successful if performed at other times of the year, partic- ularly in the post-spawning period. The accuracy of the cytologic sexing of blue mussels throughout the year would require further study. It should be noted that the evaluation of unstained smears of '*4'.^4.* >?i ^?--.V- ^v.#?- .?v^; . Figure 2. Photomicrograph of squash smear (cytologic) preparation of reproductive gland tissue from a male mussel. Note spermatozoa (arrow). Wright-Giemsa slain. Magnification, 234x. Bar = 50 (jim. mantle tissue (Jabbar and Davies 1987) or direct microscopic eval- uation of mantle tissue from mussels (Nichols 1991 ) has been used previously by other investigators. However, the use of Ro- manowsky-type staining and the rigorous correlation of the cyto- Figure I. Photomicrograph of fixed tissue (histologic) section of re- productive gland tissue from a male mussel. Note spermatozoa (ar- row). Hematoxylin & eosin stain. Magnification. 234x. Bar = 50 jjim. ■f^ ■> ;i» Figure 3. Photomicrograph of fixed tissue (histologic) section of re- productive gland tissue from a female mussel. Note oocytes (arrows). Hematoxylin & eosin stain. Magnification, 234x. Bar = 50 urn. Cytologic Sexing of Marine Mussels 347 y- •-Tv^i^.v^nrx \l-' V y I ■,»*«»ao Figure 4. Photomicrograph ol squash siiuar (c\liil(ij;ii I pripaialion of reproductive gland tissue from a female mussel. Note oocytes that have ruptured in smear preparation (arrows). Wright-Giemsa stain. Magnification, 234x. Bar = 50 jim. the ease of sample preparation and evaluation, the low cost of slides and stain, short time (approximately 10 min for drying and staining), and minimal tissue required. Although this study was initiated to answer producer concerns related to marketing, this procedure may be of use to researchers interested in determining if sex influences the biochemical content of tissues, growth, or phys- iologic parameters. Although the sexing in this study was a ter- minal procedure (to attain sufficient samples for histologic pro- cessing), cytologic sexing requires very minimal tissue and could possibly be attained by needle aspiration, as is routinely performed in veterinary and human medicine. Fine-needle aspirates of mantle tissue (using a 21-gauge needle and 6-ml syringe) may result in diagnostic cytologic preparations with minimal effect on the mus- sel, allowing the cytologic sexing of live mussels in which further evaluations requiring living mussels were to be performed. Ade- quate preparations were achieved by us when doing fine-needle .ispirates of mantle tissues (data not reported), but this technique was not consistently evaluated in this study. In summary, the cytologic evaluation of squash preparations of mussel tissue was highly accurate in determining the sex of marine mussels (M. ediilis) when evaluated in late April in Prince Edward Island. Canada. The advantages of the cytologic procedure over traditional histologic processing and biochemical determination include ease of sample preparation and evaluation, short prepara- tion time, low cost, and the spanng of tissue for other studies. ACKNOWLEDGMENTS logic method to the traditional standard of histologic sexing are features that distinguish this study from earlier ones. The advantages of the cytologic sexing of marine mussels are The authors thank Mr. Allan Mackenzie, Mr. Neil MacNair, and Ms. Isabclle Dutil for assistance in mussel processing, as well as Mr. Brian Fortune for providing the mussels. LITERATURE CITED Jabbar. & J. 1. Davies. 1987. A simple and convenient biochemical method for sex identification in the marine mussel. Mxnhis ediilis L. J . Exp. Mar. Biol. Ecol. 107:39^4. Luna. L. G. 1968. Routine staining procedures, pp. Al-Afj. In: L. G. Luna (ed.) Manual of Histologic Staining Methods of the Armed Forces Institute of Pathology. McGraw-Hill. Toronto. Martin, S. W., A. H. Meek & P. Willeburg (eds.). 1987. Sampling meth- ods, pp. 22-47. In: Veterinary Epidemiology. Principles and Methods, Iowa State University Press. Ames. Iowa. Nichols, S. J. 1991, Determining the sex and reproductive status of zebra mussels, p, 51. Proceedings from the 2nd International Zebra Mussel Research Conference, November 19-22, 1991. Tyler. R D . R, L, Cowell. C. G. MacAllister & R. J. Morton. 1992. Introduction, pp, 1-20. In: R. L. Cowell and R. D. Tyler (eds.|. Cy- tology and Hematology of the Horse. American Vetennary Publica- tions. Goleta. California. Journal of Shellfish Rfifiirch. Vol. 15. No. 2. 349-353. 1996. COMPARATIVE ALLOMETRIES IN GROWTH AND CHEMICAL COMPOSITION OF MUSSEL (MYTILUS GALLOPROVINCIAUS Lmk) CULTURED IN TWO ZONES IN THE RIA SADA (GALICIA, NW SPAIN) M. JOSE FERNANDEZ-REIRIZ, UXIO LABARTA, AND JOSE M. F. BABARRO C.S.I.C. Inslitiito de Investigaciones Marinas Ediuvdo Cabello 6. 36208 Vigo. Spain ABSTRACT A study was carried out in two zones in the ria de Sada (Galicia, NW Spain) with mussels (Mytiliis galloprovincialis Lmkl coming from the same stock and cultured in rafts, from seed sized to the first split (individuals with sizes between 15 and 65 mml. in order to evaluate a) the environmental differences in each location and b) their influence on growth (length and dry weight) and on chemical composition (protein, carbohydrates, glycogen, lipid classes, and fatly acids of total lipids). Allometric relations regarding those parameters for mussels on both locations were also determined. The mussels cultured in the inner zone (Amela), with a higher concentration in chlorophyll a dunng the period studied and a smaller number of cultured mussels, present the highest performance in dry weight and a nearly doubled amount of glycogen than those in the outer one (Lorbe). There are also significant differences in saturated, monounsaturated. nonmethylene-interrupted dienoic. and total fatty acids and sterols, the lowest content being shown by those cultured at Amela. The mussel is an important dietetic source of polyunsaturated iu-3 fatty acids, as eicosapenlaenoic acid and docosahexaenoic acid; both of them are beneficial for health (i.e.. the treatment and prevention of cardiac ischemia). The values found range from I to 2% of dry weight in both locations. An indicator of the greatest importance if the product is to be used for human food is the ratio of ui-Slui-i fatty acids. Habitual diets with a ratio of (o-6 to (u-3 fatty acids of 12-50 are associated with a greater risk. The value observed in this study was about 0.2 in both locations. A'£y' WORDS: Mussel, growth, chemical composition, dietetic value INTRODUCTION The growth rate of bivalve molluscs is mainly determined by the environmental conditions in the zone of culture. Different en- vironmental factors- (temperature, salinity, primary production, and food availability among others) have a clear influence on their growth (Perez Camacho et al. 1991). Zones with different lati- tudes, with few climatic changes (northern and southern Galician bays), present, northwards, a delayed gradient in the sexual mat- uration period (Villalba 1995). Taking this as an aim, a study has been carried out on two zones of the ria de Sada (Galicia, NW Spain) with mussels {Mylilus galloprovincialis Lmk.) coming from the same stock and cultured in rafts, from seed size to the first split, comprising the period from winter to spring (individuals with sizes between 15 and 65 mm), in order to evaluate a) the environmental differences in each location and their influence on growth (length and dry weight) and b) the biochemical composition (protein, carbohy- drates, glycogen, lipid classes, and fatty acids of total lipids). Another one of the authors" objectives is the reassessment of mussel in order to improve its consumption. Heam et al. (1989). Ackman (1990), and Cronin et al. (1991). among others, have clearly demonstrated that polyunsaturated a)-3 fatty acids (o)- 3PUFA) help diminish both the risk of cardiovascular illnesses and the level of tryglicerides and can increase the level of high-density lipoproteins. Concerning to-6 fatty acids (co-6PUFA), these au- thors say that they have strong effects on humans; however, some of these effects are antagonists to the ones proposed for the aj-3PUFA. This explains the importance of studying the biochem- ical characteristics of organisms that will be used as human food. MATERIALS AND METHODS Culture Conditions Ropes with mussel seeds from collectors (seeds fixed in the ri'a) and with an initial average size (length) of 27.61 ± 7.28 mm were placed in two rafts in the ria de Sada (Galicia. NW Spain), one at the inner zone (Amela) with an area of 490,899 m" on 30 rafts and another at the outer one (Lorbe) with 1,339,200 m" on 90 rafts. The density estimated for mussel culture was 89.5 kg/m^ at Lorbe and 43.1 kg/m' at Amela. The initial sample was carried out in November 1992 (size frequency and dry weight by size), together with a subsampling of the different size groups, to perform the analyses conceming bio- chemical composition. The final sampling at each of the rafts was performed in April 1993. Temperature and chlorophyll a in the water column were made by means of CTD (Sea Bird 25) cast with a fluorometer (Sea Teach). The fluorescence values were calibrated against acetone extracts of chlorophyll a (Yentsch and Menzel 1963). The chlo- rophyll a was expressed in milligrams per cubic meter. Condition index was calculated according to Davenport and Chen (1987). Biochemical Composition A mean of 50 mussels for the lower class of size (15-35 mm) and 25 mussels for the upper class of size (40-65 mm) were freeze dried and stored at — 70°C under inert nitrogen atmosphere for later analysis. The samples were sprayed in at pulverisette 6 (Fritsch) and homogenized with water in an ultrasonic vibrator Sonifier 250. Proteins were studied following the method de- scribed by Lowry et al. ( 195 1 ), after hydrolysis with NaOH 0.5 N for 24 h at 30°C. Carbohydrates were quantified as glucose by the phenol-sulphuric acid method (Stickland and Parsons 1968). Gly- cogen is also quantified as glucose after precipitation with 1(X)% ethanol. Lipids were extracted following a modification of Bligh and Dyer (1959), described by Fernandez Reiriz et al. (1989). Fatty acids from total lipids were transesterified to methyl esters with HCl in methanol and later analyzed by gas chromatography as descnbed by Fernandez Reiriz et al. (1993). Nonadecanoic acid 349 350 Fernandez-Reiriz et al. was used as an internal standard, and a response factor was cal- culated for each fatty acid in order to perform quantitative analy- ses. Lipid classes were studied as described by Fernandez Reiriz et al. (1993). RESULTS The two zones studied are located in the north area of the ria de Sada (Galicia, NW Spain) (Fig. I), 2 miles from each other. During the period studied, winter rainy weather, temperatures were similar (12.83 ± 0.64°C at Lorbe and 12.92 ± 0.62°C at Amela). Concerning chlorophyll a. the change during the period studied is shown in Figure 2. The mean values at each station are 17.10 ± 11.63 mg/m^ for Amela and 10.66 ± 8.15 mg/m' for Lorbe. At the beginning of the experience, the mean size of the mussel seed was 27.61 ± 7.28 mm. At the end, the average size reached by the Amela mussel was 40.73 ± 0.53 mm, whereas the one cultured at Lorbe attained 34.89 ± 0.44 mm. The evolution of the condition index (Table 1 ) under size groups at both locations shows significant differences from size group 50 on. AUometric relations (y = ax^) weight/size from these data respond to the following equation: Amela y = 0.00264X-""'- (r = 0.996, p < 0.0001, n = 10) Lorbe y = 0.00668x-^" (r = 0.998. p < 0.0001, n = 10) The slopes of the allometric relation between locations were compared by means of a covariance analysis, after logarithmic R\A otS f^OA Figure I. Location in the ria de Sada (Galicia, NW Spain) of the two zones studied. I I Amela Lorbe i^' — CM ro — CM CM 1992 I 1993 Figure 2. Evolution of chlorophyll a imglm') during the period stud- ied. TABLE 1. Condition index of size groups at both locations. Size of Class Arnela Lorbe 15 20 25 30 35 40 45 50 55 60 10.0 12.0 11.9 14.7 17.3 18.9 20.4 22.3 23.6 25.8 8.7 12.2 12.3 15.8 17.3 18.6 16.9 19.7 19.6 18.5 transformation. The analysis shows significant differences (Fl,16 = 6.12, p < 0.025) between slopes, depending on the zone of culture. The highest performances in dry weight flesh were at- tained at Arnela. Chemical Composition (Protein. Carbohydrates, Glycogen, Total Lipids, Lipid Classes, and Fatty Acids) The relations among the different chemical parameters (y, mil- ligrams) and the size (x, millimeters) in both locations are de- scribed by means of the following equations; Prolein: (r = 0.986, p < 0.0001, n = 10) (r = 0.998, p < 0.0001, n = 10) Amela Lorbe 0. 00276 x-***" 0. 00340 x-'*-'° Carbohydrates: Amela Lorbe Glycogen: Arnela Lorbe Lipids: Arnela Lorbe = 0.00009 x-^-*'-^ = 0.000I9X' '-"' = 0.00008 x'-" = 0.00005 x'^"' = 0.00021 x'-'* = 0. 00050 x-'^" (r = 0.973, p < 0.0001, n = 10) (r = 0.988, p < 0.0001, n = 10) (r = 0.946, p < 0.0001, n = 10) (r = 0.958, p < 0.0001, n = 10) (r = 0.980, p < 0.0001 (r = 0.994, p < 0.0001 10) 10) The slopes of the allometric relations between the different Allometries of Mussels Cultured in Two Zones in NW Spain 351 locations were compared by a covariance analysis. According to it, there are no significant differences (p > 0.05) between slopes. Regarding intercept, there are only significant differences in gly- cogen (Fl,16 = 14.17. p < 0.0025). These results confirm that the relationship of glycogen content/size was the same for both Amela and Lorbe mussels, although glycogen content at Amela is nearly double that for Lorbe mussels. If the initial size (27.61 mm) and the standard final one (stated as 40 mm) are inserted in the allometric equations of the initial sampling and in the final (organic componentVsize) one. several changes in the different components studied can be observed (Ta- ble 2). Regarding the initial mussel, total increments in protein, carbohydrates, glycogen, and lipids are, in general, higher in the mussel presenting a greater growth rate, so these increments de- pend on the global growth of mussel. Worth noting is that protein continues to be the main component in both locations and that Amela mussels have twice as much glycogen as those from Lorbe. Lipid Classes Relations among the different lipid classes (y. milligrams) and size (X. millimeters) are described by the following regressions; Phospholipids: Amela y = 0.00007x'«' Lorbe y = O.OOOOSx'™' Triacylglycerols: Amela y = 0.00008 x-^*- Lorbe y = 0.00021 x-"^' Sterols: Amela Lorbe y = 0.00001 X 3 °>5 y = 0.00005 x-<^" 0.998. p < 0.0001 0.999. p < 0.0001 10) 10) 0.842, p < 0.0014, n = 10) 0.889, p < 0.0003. n = 10) 0.989. p < 0.0001, n = 10) 0.998. p < 0.0001. n = 10) Covariance analyses identified only significant differences be- tween slopes for sterols (Fl,16 = 5.45, p < 0.05). At the end of the research, phospholipids became the main lipid class (Table 2). There are no significant differences in phospholipidsand tryglic- erides between the locations. Fatty Acids Fatty acids content (y, milligrams) varies with size (x, milli- meters), according to the following allometric equations: TABLE 2. Biochemical changes (expressed as milligrams per individual) during the culture of mussel {M. galloprovincialis Lmk), Initial Final Component Arnela Lorbe Protein 39.78 120.80 116.22 Carbohydrate 6.23 35.75 28.69 Glycogen 3.93 13.34 7.64 Lipids 5.87 28.78 28.05 Lipid classes Phospholipids 1.18 15.92 16.19 Triacylglycerols 1.25 3.34 4.06 Sterols 1.35 0.80 0.87 Fatty acids Saturated 0.77 5.00 6.61 Monounsalurated 0.34 2.34 3.35 PUFAs 0.57 5.75 6.39 M-3PUFAS 0.31 4.97 4.75 CU-6PUFAS 0.10 2.11 0.73 NMID 0.05 0.19 0.35 Saturated: Arnela y = 0.000009 x^"" Lorbe y = 0.000065 x"" Monounsaturated: Arnela y = 0.000013x^^82 Lorbe y = 0.000037 x ' ""^ Polyunsaturated: Amela y = 0.000018 x ' •»3'' Lorbe y = 0.000042x ' -"* Arnela y Lorbe y Sco-7.- Arnela y = 0.000002 x-^ = 0.000006 x""-' = 0.000008 x'""-' Lorbe Amela Lorbe 0.000002 X (r = (r = (r = (r = (r = (r = (r = (r = (r = (r = 0.00001 ix'--' (r 0.000037 x-"'' (r ixi-3PUFA: Arnela y = 0.000018x' 3''' (r = Lorbe y = 0.000021 x'^''- (r = Noitmethylene-interrupted dienoic Arnela y = 0.00001 1 x- "" (r = Lorbe y = 0.000024 x-^'J" (^ = 0.991, p < 0.0001, n = 10) 0.998, p < 0.0001. n = 10) 0.993. p < 0.0001. n = 10) 0.984. p < 0.0001. n = 10) 0.988. p < 0.0001. n = 10) 0.996, p < 0.0001, n = 10) 0.991, p < 0.0001, n = 10) 0.985, p < 0.0001, n = 10) 0.964, p < 0.0001, n = 10) 0.982, p < 0.0001, n = 10) 0.996. p < 0.0001. n = 10) 0.996, p < 0.0001, n = 10) 0.981, p < 0.0001, n = 10) 0.996, p < 0.0001, n = 10) (NMID): 0.983, p < 0.0001, n = 10) 0.995, p < 0.0001, n = 10) 0.989, p < 0.0001, n = 10) 0.997, p < 0.0001, n = 10) 0.931, p < 0.001, n = 10) 0.980, p < 0.0001, n = 10) The covariance analysis carried out to compare the slopes of the allometric relations between locations shows that there are only significant differences between slopes for PUFA (Fl .16 = 12.51, p< 0.005), for Sa)-9(F1, 16 = 5.51, p < 0.05), and for the total of fatty acids (Fl,16 = 18.9, p < 0.001). As for intercepts, differences were only detected for monounsaturated fatty acids (F1.16 = 34.82. p < 0.0005), NMID (F1,I6 = 18.9, p < 0.001), and the (D-3/a)-6 relation (Fl,16 = 11.96, p < 0.005). Therefore, the relation between content of monounsaturated and NMID fatty acids and size was the same for both Arnela and Lorbe mussels, although the content in those fatty acids was higher at Lorbe. Saturated and polyunsaturated (PUFAs) fatty acids are the main groups. Monounsaturated ones form a smaller group (Table 2). In our study, we have found that M. galloprovincialis has high con- tents of essential oj-3 PUFAs and low levels of NMID. DISCUSSION The reserve cycles of mussel indicate a complex interaction between food and temperature and between growth and the game- togenic cycle. The association between gonadal development and the reserve accumulation cycles is well known (Gabbott 1983). Growth and Biochemical Composition In our study with similar temperatures in both locations, Amela presents a significantly higher (p < 0.05) concentration of chlo- rophyll a than does Lorbe, so the slower growth shown by Lorbe Total fatty acids: Amela y = 0.000041 x' 445 (r Lorbe y = 0.000140x^ 163 (r Relation a )-J/co-6; Amela y = 1.59x"-"^ (r Lorbe y = 1.31 x"^-^- (r 352 Fernandez-Reiriz et al. mussels can be explained by the smaller availability of food and chlorophyll a at this location. These differences in primary pro- duction offer different growth rates between mussels at both lo- calities. Perez Camacho and Roman (1979). Dickie et al. (19841. and Mallet and Carver (1989) clearly state the influence that the zone of culture has on Mylihis ediilis growth. Rodhouse et al. (1984) indicate that in many situations, food availability is the most im- portant individual factor acting on mussel growth. Following this trend. Page and Hubbard ( 1987) detect a clear relation between the concentration of chlorophyll <; and the growth in length of M. edulis. They did not find any connection between this growth and water temperature. While studying the gametogenic cycle of mussels in the bays of Galicia. Villalba (1995) observed a slowed down gonadal devel- opment in Lorbe mussels, together with small growth rates, with regard to other bays of Galicia. Gabbot ( 1976) observed that the use of glycogen in M. edulis is closely related to its gametogenic cycle. Therefore, the highly significant contents of glycogen of Amela mussels found in this study could be related to the game- togenic cycle, which is directly related to the available food in each locality. Further study of the gametogenic cycle of mussels from both areas could confirm this hypothesis. Lipid Classes The highest contents of phospholipids and triacylglycerols were observed in April in both locations. Beninger and Lucas (1984) observed in Tapes dectissaliis and in Tapes phiiippinanim a similar variation for phospholipids and triglicerides. with the highest values in spring (April). For those authors, these values are in relation to the active gametogenesis period. In our study, we only have data at the beginning (November) and at the end (April) and it only can be imputed to available food cycles in the "rias" and energetic reserve utilization for the mussels but without evi- dence of seasonal changes. There is little information on sterol changes, apart from their having an unespecified role in gonadal development (Gabbott 1976). Fatty Acids It is worth noting the huge contribution of a)-3PUFA (20:5(d-3 and 22:6to-3) to the total of fatty acids, representing about 29 and 38% of them for Lorbe and Amela, respectively. A fact also described by Gabbott (1976). Chu et al. (1990) say that the vari- ations observed in co-3PUFAs can be due both to the diet and to the reproductive cycle. Pollero et al. (1979) in Crassostrea tehuelcha and Trider and Castell (1980) in Crassostrea virginica studied the role of C0-3PUFA and observed that the level of 20;5to-3 increased before the hatching and decreased afterwards. With regard to NMID, and although their function is not clearly discerned, it seems that their biosynthesis is regulated both func- tionally and physiologically. Actually, big amounts of these acids were observed in the membranes of sponges (Morales and Lich- field 1986) and molluscs (Ackman and Hooper 1973). which would indicate a functional and structural role in the biological system. Klingensmith (1982) cleariy indicates a competitive in- corporation between NMID and PUFAs, especially 20;5to-3 and 22:6a)-3. Fang et al. (1993). working with deep-water mussels in the Gulf of Mexico, found that they had a high NMID content, but not essential PUFAs. In our study, we have found M. gallopro- vincialis to have high contents of essential PUFAs and low levels of NMID. With regard to the increase of saturated and monoun- saturated fatty acids, Waldock and Holland (1979). suggest that this increase is the result of the synthesis de novo from glycogen. Dietetic Value Fatty acid composition in marine organisms is a topic of great interest these days, because of the beneficial effects co-3PUFA seem to have in order to reduce deaths caused by cardiovascular illnesses. Although the ideal amount a human diet must take in order to prevent them is not yet known, the effect its consumption produces has been already stated. Ackman ( 1990). following We- ber, shows the existence of a positive correlation between deaths by coronary illnesses and a high relation of u)-6/io-3 (between 12 and 50). so the effect that lipid composition has on the quality of a product destined for human consumption is important. Mussels studied here have a high content of to-3PUFA. between 29 and 38% of the total of fatty acids (1 and 2% dry weight), and a (D-6/to-3 relation of 0.2 in all size groups studied, a much lower one than those cited as harmful in the literature. CONCLUSIONS In big firms of mussel culture, management departments call for detailed knowledge about culture areas, their environmental factors, biological cycles, and the final characteristics of the prod- ucts they are to obtain, in order to make the most of production and marketing. This work has been done taking all of these demands into account, and the following conclusions have been drawn: 1 . Mussels cultured at Amela present higher growth rates and also a higher condition index than those cultured at Lorbe; this is clearly related to food availability and the density of culture in each area. 2. The chemical composition of the mussels of the two zones showed a similar pattem, although some significant differ- ences were observed in the contents of certain components: Amela mussels present the highest contents of glycogen, as well as the lowest ones in saturated, monounsaturated. and NMID fatty acids and sterols. These facts show an influence of the habitat in the biochemical components of mussels, beyond those probably conditioned by the gametogenic cy- cle. 3. As far as the dietetic value of this product is concerned, it is important to note the high content of co-3PUFA (between I and 2% dry weight) found in the mussels of both locations and the low io-6/a)-3 relation (about 0.2). ACKNOWLEDGMENT We thank J. L. Garrido. Lourdes Nieto. Beatriz Gonzalez, and Ana Ayala for their help with the analyses and Amalia Collazo for the translation. This work was supported by the contract-project with the Conselleria de Pesca de la Xunta de Galicia. the firm PROINSA. for supplying biological material and giving every chance for sampling. LITERATURE CITED Ackman. R. G. 1990. Sea food lipids and fatty acids. Food Rev. Int. with the sand shrimp (Cningon septemspinosa) . Camp. Biochem. 6(4):617-646. Physiol. 64B:15_VI65. Ackman, R. G. & S. N. Hooper. 1973. Non-methylene interrupted fatty Beninger, P. G. & A. Lucas. 1984. Seasonal variations in condition. acids in lipids of shallow- wat {Litorina litorea and Lunatia triseriata) reproductive activity, and gross biochemical composition of two spe- Allometries of Mussels Cultured in Two Zones in NW Spain 353 cies of adull clam reared in a common hahitat; Tapes decussaliis L (Jeffreys) and Tapes philippmarum (Adams & Reevel. J . Exp. Mar. Biol. Ecol. 79:19-37. Bligh. E. G. & W. J. Dyer. 1959. A rapid method of total lipid extraction and punfication. Can. J. Biochem 37:911-915. Chu Fu-Lin. E., K. L. Webb & J. Chen. 1990. Seasonal changes of lipids and fatly acids in oyster tissues (Crassostrea virginica) and estuanne particulate matter. Comp. Biochem. Physiol. 956:385-391. Cronin. D. A.. R Powel & R. Gormley. 1991. An examination of the a)-3 and uj-6 polyunsaturated fatty acid status of wild and fanned atlantic Salmo iSalmo solar). Irish J Food Sci. Technol. 15:53-62. Davenport, J. & X. Chen. 1987 A comparison of method for the assess- ment of condition in the mussel [Mylilus ediilis L.). J. Moll. Snidies 53:293-297. Dickie, L. M., P. R. Boudreau & K. R. Freeman 1984. Influences of stock and site on growth and mortality in the blue mussel {Mylilus edulis). Can. J. Fish Aquat. Sci. 41:134-140. Fang, J., P. A. Comet. J. M. Brooks & T. L. Wade. 1993. Nonmethy- lene-interrupted fatty acids of hydrocarbon seep mussels: occurrence and significance. Comp. Biochem Physiol. 1046:287-291. Fernandez Reiriz, M. J.. A. Perez Camacho. M J Ferreiro, J. 6lanco, M. Planas. M. J. Campos & U. Labarta. 1989. Biomass production and variation on the biochemical profile (total protein, carbohydrates. RNA, lipids and fatty acids) of 7 species of microalgae. Aquacultura 83:17-37. Fernandez Reiriz. M. J., U. Labarta, & M. J. Ferreiro. 1993. Effects of commercial enrichment diets on the nutritional value of the rotifer (Brachionus plicatilis). Aquaculture 112:195-206. Gabbott. P. A. 1976. Energy metabolism, pp. 293-355. In: 6. L. 6ayne (ed.). Marine Mussels: Their Ecology and Physiology. Cambridge University Press. Cambndge. Gabbott. P. A. 1983. Developmental and seasonal metabolic activities in manne molluscs, pp. 165-217. In: A. S. M, Saleuddm and K. M. Wilbur (eds). the Mollusca vol. 2. Academic Press, New York. Heam, T. L., A. S. Sandra & S. S. Demetrios. 1989. Polyunsaturated fatty acids and fat in fish flesh for selecting species for health benefits. J. FoodSci. 52:1209-1211. Klingensmith, J. S. 1982. Distribution of methylene and nonmethylene- interrupted dienoic fatty acids in polar lipids and tnacylglycerols of selected tissues of the harshell clam (Mercenana mercenana). Lipids 17:976-981. Lowry, O H., N. J Rosebrough & A. J. Pharr. 1951. Protein measure- ment with the Folin phenol reagent. J. Biol. Chem. 193:265-275. Mallet. A. L. & C. E. Carver. 1989. Growth, mortality and secondary production in natural populations of the blue mussel. Mylilus edulis. Can. J. Aquat. Sci. 46:1 154-1 159. Morales. P. W. & C. Lichfield. 1986. Unusual C24, C25, C26 and C27 polyunsaturated fatty acids of the marine sponge Microciona prolifera. Biochim. Biophys. Acta 431:206-216. Page. H. M. & D. M. Hubbard 1987. Temporal and spatial patterns of growth in mussels Mvtilus edulis on an offshore platform: Relation- ships to water temperature and food availability. J. E.xp. Mar. Biol. Ecol. 111:159-179. Perez Camacho, A. & G. Roman. 1979. Estudio del mejillon, de su epifauna en los cultivos flotantes de la ria de Arosa. II Crecimiento, mortalidad y produccion del mejillon. Bol Inst. Esp. Oceanografia 5(l):21-42. Perez Camacho. A.. R. Gonzalez & J. Fuentes. 1991. Mussel culture in Galicia (NW Spain). Aquaculture 94(2-3):263-278. Pollero. R. J.. M. E. Re & R. R. Brenner. 1979. Seasonal changes of the lipids of the mollusc Chlamys theuuelcha. Comp. Biochem. Physiol. 64A:257-263. Rodhouse. P. G.. C. M. Roden. G. M. Bumell. M. P. Hensey. T. Mc- Mahon. 6. Ottaway & T. H. Ryan. 1984. Food resource, gametoge- nesis and growth of Mytilus edidis on the shore and in suspended culture: Killary Harbour. Ireland. J. Mar. Biol. Assoc. U.K. 64:513- 529. Stnckland. J. D. & T. R. Parsons. 1968. A practical handbook of sea water analysis. Bull. Fish. Res. Board. Can. 167:173-174. Trider. D. J. & J. D. Castell. 1980. Influence of neutral lipid on seasonal variation of total lipid in oysters. Crassostrea virginica. Proc. Natl. Shellfish. Assoc. 70:112-118. Villalba, A. 1995. Gametogenic cycle of cultured mussel. Mytilus gallo- prorincialis. in the bays of Galicia (NW Spain). Aquaculture 130:269- 277. Waldock. M. J. & L. Holland. 1979. Seasonal changes in the triacylglyc- erol fatty acids of the mantle tissue of the mussel Mytilus edulis L. Biochem. Soc. Trans. 7:898-900. Yentsch, C. S. & D. W. Menzel. 1963. A method for determination of phytoplankton chlorophyll and phaeophytin by fluorescence. Deep Sea Res. 10:221-231. Journal of Shellfish Research. Vol. 15, No. 2, 355-362. 1996. SIZE-DEPENDENT SURVIVORSHIP OF THE BIVALVE YOLDIA NOTABILIS (YOKOYAMA, 1920): THE EFFECT OF CRAB PREDATION MASAHIRO NAKAOKA Ocean Research Institute University of Tokyo Minamidai 1-15-1 Nakano, Tokyo 164. Japan ABSTRACT Age-specific .survivorship of the bivalve YoUiia iwiahilis was estimated by quantitative field samplings, and the possible effect of predation by the crab Paradorippe gramilala on its survivorship pattern was examined by laboratory experiments. The annual monality rate in the field was high (>86'7r) in the two youngest year classes, whereas mortality was lower (<40'7f ) in the classes older than 3 y. leading to Deevey's type 3 survivorship curve. Field mortality was size dependent; bivalves smaller than 10 mm in shell length suffered high mortality ( >68'^). whereas those larger than 10 mm did not ( <40'7t ). regardless of actual age. Burrowing depth of Y nolahilis increased with shell length, although considerable variation was found among larger individuals. The laboratory expenment revealed that crabs preferred smaller bivalves The resultant size-related mortality pattern in the experiment was similar to that in the field, suggesting that size-selective predation by P. granulata contributes to the size-specific survivorship pattern of Y. nolahilis. Size-selective predation by this crab did not depend on the size-dependent burrowing ability of the bivalve, because small sizes were preferred even when bivalves of all sizes were constrained to live in shallow sediments. KEY WORDS: Yotdia. survivorship, crab, size-selective predation. burrowing depth, laboratory experiment INTRODUCTION For understanding the population dynamics of organisms with a long hfespan, it is essential to estimate age- or size-specific changes in survivorship. Age-specific survivorship has been de- termined for several marine bivalves, including some long-lived species such as Chlamys islandica (Vahl 1981), Macoma bcdthica (Green 1973). Mercenaha mercenaria (Kennish 1980), Mya are- naria (Brousseau 1978, Goshinia 1982). Mytilus edidis (Thomp- son 1984), and (7,s7^«i c(/h/;,v (Walne 1961). Most of these animals show high mortality as juveniles and low mortality as adults, which leads to an negative exponential survivorship cui^e, or the type 3 curve of Deevey (1947). Size-dependent mortality factors are often responsible for producing this survivorship pattern. Size- dependent mortality may result from the relation of size to preda- tion (Paine 1976, Seed 1993) or competition (Bell and Coull 1980) or differences in vulnerability to physical stress or disturbance (Levinton and Bambach 1969. Hughes 1970. Rhoads and Young 1970, Woodin and Marinelli 1991). Size-selective predation by predators such as benthic fish, crabs, and sea birds has been well established as setting the sur- vivorship patterns of marine bivalve populations (Paine 1976. Seed and Brown 1978. Holland et al. 1980. Peterson 1982a. Peter- son 1982b. Arnold 1984. Sanchez-Salazar et al. 1987. Zwarts and Blomert 1992b, Seed 1993). For infaunal bivalves, the effect of predation is often affected by the burrowing depth of individuals (Hughes 1970, Blundon and Kennedy 1982, Moller and Rosen- berg 1983. Elmgren et al. 1986. Haddon et al. 1987. Zwarts and Blomert 1992a). Field observations and experiments have been undertaken to test the possible effects of predators on intertidal and shallow subtidal bivalve populations, but rarely have such studies been conducted at water depths deeper than 10 m. mainly because of difficulties in estimating field mortality, as well as logistical constraints in carrying out field experiments. Yoldia notabilis is a deposit-feeding bivalve that lives in muddy-sand bottoms of the northwestern coast of the Pacific. It occurs at high densities (ca. 200 m~~) at depths of 10-15 m in Otsuchi Bay. northeastern Japan. Age is easily determined by counting the number of external and internal shell growth rings (Nakaoka 1992a. Nakaoka and Matsui 1994). It is therefore pos- sible to quantify its age-specific survivorship by monitoring the density of each year class over time. Yoldia spp. are often con- sumed by benthic predators such as fish and crabs (Tyler 1972. Jewett and Feder 1980). In Otsuchi Bay. the distribution of Y. mnabilis overlaps that of potential epibenthic predators such as the crabs. Paradorippe granidala and Ovalipes punctatus and the flounders Paralichtys oUvaceus and Kareiiis bicoloratus. P. gran- ulata. in particular, is known to be an active bivalve feeder (Sasaki 1993. Goshima personal communication). The predation pressure from these predators may affect the age- and size-specific survi- vorship of Y. notabilis. The main objectives of this article are ( I ) to test whether the survivorship of the field population of Y. notabilis is dependent on size, and (2) to evaluate the potential for size-selective predation by P. granidaia to contribute to the observed survivorship pattern of }'. notabilis. 1 carried out a field census to quantify densities of the bivalve and the crab and also laboratory predation experiments using these animals. 1 also tested whether size-selective predation by the crab depends on the burrowing depth of the bivalve. MATERIALS AND METHODS Estimation of Field Densities A field census was carried out between June 1990 and June 1991 at two stations (Stn YA. 10 m in depth; Stn YD, 14 m) established at the inner part of Otsuchi Bay (see map of sites in Nakaoka 1992b). Four to 10 quantitative sediment samples were taken monthly at each station with a 0. 1-m" Smith-Mclntyre grab sampler. The collected sediments were sieved through a 1-mm- mesh sieve, and individuals of Y. notabilis and P. granulata re- tained on the sieve were sorted out. The age of each individual Y. notabilis was determined by counting the annual growth rings on the external shell surfaces. A previous study has shown that these rings are formed annually during winter (Nakaoka 1992a). Indi- viduals were classified into one of seven year classes between 355 356 Nakaoka Class 1983 (8 y old in 1991) and Class 1990 (1 y old) or a compound class that consisted of individuals that were recruited before 1983 (Class <1983; >8 y old). Nakaoka (1992a) has shown that the youngest year class (Class 199 1 : 0 y old) cannot be collected by this method because the individuals are so small that they pass through the 1-nim-mesh. To estimate the density of this class, three to four subsamples each of 0.01 m" were taken from the grab samples (one subsample from each grab sample) monthly at Stn YD (except March 1991) and on six occasions (June, July, and November 1990 and April, May. and June 1991) at Stn YA, and they were sieved through a 0.5-mm-mesh sieve to collect 0-y-old individuals. For all of the bivalves, shell length was mea- sured with a caliper to the nearest 0. 1 mm and the density of each year class was determined for each month and station. The number of P. gramdata collected by the same grab sampler was also recorded. More detailed information on the study sites and sam- pling procedure is given in Nakaoka (1992b). Burrowing Depth The in situ burrowing depth of Y. notabilis was measured from grab samples at Stn YD on January 22, 1990. Six 8-cm-deep sediment cores (each 0.01 m" in area) were taken from three grab samples, and the cores were immediately divided on shipboard into 2-cm sections. The sliced sediments were sieved through a 1-mm-mesh. Individuals of Y. notabilis in each section were counted, and the shell length of each individual was measured. Because the sediments deeper than 8 cm were not taken quan- titatively, the burrowing depth was also observed in the labora- tory. Individuals were collected at the same date as above and kept alive in the laboratory in two aquaria (each 0.01 m'^ in area). Each aquarium contained 15 individuals of sizes ranging between 4.2 and 38.4 mm in shell length. Aquaria were filled to a depth of 10 cm, with sediment collected at Stn YD, and supplied with running seawater. Four days later, the sediments in the aquaria were taken out at 2-cm intervals to determine the burrowing depth of each individual. Experiment 1 : Size-Specific Crab Predation on Bivalves A laboratory exjjeriment testing for the size-selective predation by P. granulata on Y. notabilis was carried out from July II to August 8, 1992. Individuals of Y. notabilis and P. granulata were collected with the grab sampler and a biological dredge with a 50-cm mouth opening at Stn YA and Stn YD a day before the experiment. Eighteen individuals of Y. notabilis (2.9-39. 1 mm in shell length) were held in each of six aquaria (each 0.025 m" in area). The size distribution of the bivalve was similar in all aquaria. Four aquaria had one crab (16.7-20.6 mm in carapace length) added to each of them (experimentals), and the remaining two did not (controls). The resulting densities of Y. notabilis and P. granulata (720 and 40 m~", respectively) were greater than those observed in the field (179 and 206 m " at Stn YA and YD, respectively, for Y. notabilis: 1.8 and 1.7 m~' for P. granulata: see Results; see also Nakaoka 1992b) because of the limitation of experimental space. The sediment used for the experiment was taken from Stn YA and Stn YD, sieved through a I-mm-mesh sieve, mixed, and dried under sunlight for more than 3 d to remove any live animals. The aquaria were filled with this sediment to a depth of 8 cm, and running seawater was supplied to each aquar- ium. Water temperature in the aquaria varied no more than 2°C from that of the natural environment ( 15.9-19.0°C during the ex- perimental period; Takagi et al. 1992). I first introduced bivalves to the aquaria and then crabs on the following day, after all bi- valves had burrowed under the sediment surface. No alternative food was supplied to the crabs during the experiment. At the end of the experiment, sediments were removed from each aquarium at 2-cm intervals from the surface, and the sizes of all live and dead specimens of Y. notabilis in each layer were recorded. Observations of the feeding behavior of P. granulata in the laboratory revealed that crabs find Y. notabilis by scooping the sediment surface (at least to more than 1 cm deep) with their chelipeds. Crabs feed on the soft parts of the bivalve by inserting the chelipeds through a gape between two valves of the prey. In most cases, the shells are broken by this activity. Even if the shells sometimes remain unbroken, the mortality due to crab predation is distinguishable from other mortality causes in the laboratory be- cause soft tissue is completely removed from the shells by preda- tion. I classified the fate of each bivalve into three categories: ( I ) alive, (2) predated (broken shell and unbroken empty shell), and (3) mortality due to other causes (unbroken shell with soft parts remaining). Experiment 2: The Effect of Burial Depth on Crab Predation To test the effect of the burrowing depth of Y. notabilis on susceptibility to P. granulata predation, I changed the sediment depth in the aquaria in the second laboratory experiment (July 8-August 14, 1993). I used a methodology and experimental pro- tocol similar to those in Experiment 1 , but with a different exper- imental design. Eighteen individuals of Y. notabilis (4. 1-36.3 mm in shell length) and a crab ( 19.0-23.0 mm in caparace length) were held in each six aquaria, three containing an 8-cm layer of sedi- ment and another three with a 2-cm layer of sediment. A higher predation rate is expected with decreased sediment depth if crab predation is related to the burrowing depth of the prey (see Blun- don and Kennedy 1982, Elmgren et al. 1986 for similar experi- mental designs). No controls (aquana without the crab) were pre- pared because of an insufficient number of bivalves. Two crabs (one in each treatment) died during the experiment. The data form these two aquaria were excluded from the analysis. Data Analyses A survivorship curve for each year class of Y. notabilis at each station was obtained by transforming monthly mean density .v to logt.v -I- 1 ) and by regressing it against date /. The slope of the regression equation expresses the rate of decrease in density due to mortality. 1 tested the heterogeneity of slopes between year classes and stations using a general linear model expressed as follows; 'y* a + a,, -I- «,_, + '"lOi;!/; + P',,* + Pi,'„* + P2/,A + (gkb iP,),/ ,/ijk (I) where y,^ is the log-transformed density of year class / at station J at date t,jf.: a and p are average regression coefficients (intercept and slope, respectively); a,,, a,^. p,, and Pj/ ^e treatment effect coefficients; and (aia,),, and (P,P;),, are effects of interaction between year class and station. The F-test on the terms Pi/„^ and P^/,,;, shows whether the slope differs significantly among age classes and between stations, respectively, and the test on the term (P,P2),/,yit examines the age*station interaction effect on the slope. This test assumes that the estimate of density for each month IS independent of that for other months, which may not be true if one samples the population repeatedly at the same site. Size-Dependent Survivorship of Yoldia notabius 357 During my research, however. 1 tried not to collect sediments exactly from the same spot. The density estimates, therefore, can be regarded as independent. To test the between-site differences in age- and size-specific survivorship, the slopes of the regression equations were also com- pared between the stations for each year class and for each size class, the latter by classifying each year class into six different size categories (0-5, 10-15, 15-20. 20-25. 25-30. and >30 mm in shell length) according to the mean shell length of each year class (Nakaoka 1992b). The general linear model for this analysis was reduced to: .v„* = a -I- a„ 4- ^t„, + P,,/,,,. (2) Year class 1989 and size class 5-10 mm were not analyzed be- cause these classes did not occur at Stn YA. The annual mortality rate was calculated from the difference in predicted densities on June 1. 1990 U = 0) and June I, 1991 {i = 365), which were estimated by use of the regression equations. The relationship between bivalve size and burrowing depth was tested by the use of Fisher"s exact test of independence after cat- egorizing data into three size classes (0-10, 10-20, and 30-40 mm in shell length) and three depth layers (0-2. 2-A, and >4 cm in depth). The effect of the crab on the bivalve size-burrowing depth relationship was tested for the data of Expenment 1 by the use of a three-factor log-linear model, expressed as follows: \nf„, = [X -t- a, -t- p, + 7a + a,P, + a,7* + Py7< + ",P,7a (-^) where / is the expected frequency of a three-way contingency table; |jl is the mean of the logarithm of the expected frequencies; a,, p,, and 7^ are the effects of categories /, j. and k of three factors (treatment, size, burrowing depth); and a,P,, a,■y^. p^^^, and a,p 7i are the interaction terms expressing the dependency of two or three factors. Individual bivalves that survived until the end of the experiment were classified according to treatment (Experimen- tal and Control), shell length (0-10, 10-20, 20-30, and 30-40 mm), and burrowing depth (0-2. 2-4. and >4 cm). 1 did not test the size dependency of the crab predation rate or bivalve mortality rate in the two experiments by analysis of vari- ance or other parametric methods because the number of replicates for each treatment (« = 2-4) and the number of individuals in some size classes (see Figs. 3 and 4) were too small to obtain enough statistical power. Instead. I tested the differences in the size-frequency distribution of predated or dead bivalves using the Kolmogorov-Smimov two-sample test after pooling data for each treatment. The underiying assumption of this analysis is that in- dividuals within each aquarium behaved independently so that individuals instead of aquarium can be treated as a unit of repli- cates (see Peterson 1982b for similar analysis). This assumption may be violated in cases where larger bivalves can escape better from predators, which affects the predation rates on smaller indi- viduals. However, no observation on their behavior was made after they burrowed in the sediments. In Experiment 1, the size- frequency distributions of predated or dead bivalves in the exper- imental treatment were compared with that predicted under size- independent predation or mortality, i.e.. the initial size distribu- tion before the experiment. In Experiment 2. the difference in size-frequency distribution was tested between the two treatments. The bivalves were classified into seven size classes (0-5, 5-10, 10_15, 15-20, 20-25, 25-30, and >30 mm) in Experiment 1, but into only six size classes (4-10, 10-15, 15-20, 20-25, 25-30, and >30 mm) in Experiment 2 because of an insufficient number of the smallest individuals (<5 mm). All statistical analyses were performed with SAS (SAS Institute 1987). RESULTS Age-Specific Survivorship of the Bivalve The density of Y. nolahilis remained steady between June 1990 and June 1991 for most year classes recruited before 1988, whereas declines in density were detected in Class 1990 at Stn YA and Classes <1983 and 1987-1990 at Stn YD (Figs. I and 2). No individuals belonging to Class 1989 were collected at Stn YA during the research period. The slope of the regression lines ex- pressing survivorship curves was negative in all classes except Class 1984 at Stn YD, although the level of significance was often low (Table 1). At the 5% significance level, the slope differed significantly from 0 in five year classes at Stn YD (Classes < 1983 and 1987-1990) and in one year class (Class 1990) at Stn YA (Table 1). Younger year classes tended to have more negative values in the slope of the survivorship curve. The test of hetero- geneity of the slope revealed that the slope differed significantly between age (Table 2). The annual mortality rate was high (>86%) in Classes 1989 and 1990 and lower (<40%) in the year classes recruited before 1988 at both stations (Table 3). A large between-site difference in annual mortality was found in Class 1988. in which the annual mortality rate was 20% at Stn YA but 68% at Stn YD. A between- site comparison of slopes of the survivorship curves showed that slopes differed significantly for this year class (Table 4). The significant differences in slope were not detected in other year classes, although the estimated mortality was somewhat higher at Stn YD than at Stn YA m Classes <1983 and 1983 (Table 3). A comparison of annual mortality with mean shell length of each year class showed that annual mortality was more than 68% 10-, (a) Class <1983 ^^^i^^%-F, „ (b) Class 1983 7ih{*^'^i.4^ (c) Class 1984 E T- d c O 0 Q (d) Class 1985 6 -, m}^ \ (e) Class 1986 2 0-, (f) Class 1987 llri ^Wt^ 10 T (gl Class 1988 12 T (h| Class 1989 4 0-1 (1) Class 1990 ^4Hm- JJASONDJFMAMJJ 1990 1991 JJASONDJFMAMJJ 1990 1991 JJASONDJFMAMJJ 1990 1991 Figure I. Seasonal changes in mean density (±SE) of nine year classes of Y. notabilis at Stn YA. The individuals of Class 1989 at Stn YA did not occur during the research period. The line indicates the survivor- ship curve fitted using the parameters in Table 1. Original data and sample sizes are shown in Nakaoka (1992b). 358 Nakaoka c (1) Q (a) Class <1983 1 0 (b) Class 1983 1 0 -, i^^^Ka (d) Class 1985 6 -. UiA (e) Class 1986 2 0 (c) Class 1984 feirW^ (f) Class 1987 ^i^'Wtr 1 (1) Class 1990 .11 • i^r 1 ^^f~~^-*r- JJASONDJFMAMJJ J JASONDJFM AM JJ JJASONDJFMAM J J 1990 1991 1990 1991 1990 1991 Figure 2. Seasonal changes in mean density ( ±SE) of nine year classes of ¥. notabilis at Stn YD. See legend to Figure 1 for explanation. TABLE 2. The results of the general linear model for testing the heterogeneity of slopes of the survivorship curves in Table 1 between age classes and stations. Factor df Age Station Age*station 16.48 2.35 1.13 0.0001 0.1271 0.3490 Field Density of the Crab During the study. 1 quantitatively collected seven P. gramdata at Stn YA and nine at Stn YD, giving density estimates (average density over the research period) of 1.8 and 1.7 m"". respec- tively. The data were too few to examine for seasonal changes in density. The size range of crabs collected at Stn YA varied be- tween 15.1 and 28.1 mm in carapace length (mean, 20.5 ± 2.1 mm [SE]) and between 5.1 and 28.1 mm (mean, 18.1 ± 2.4 mm |SE1) at Stn YD. Crab sizes were not significantly different be- tween the two stations (r test; p = 0.476). for classes smaller than 10 mm in shell length and lower than 40% for those larger than 10 mm (Table 3). The test of the heteroge- neity of slopes of survivorship curves showed that the slopes did not differ significantly between stations when compared on the basis of size (Table 4). TABLE I. Parameters of the survivorship curve for each year class of Y. notabilis at the two stations. Parameters of Survivorship Curve Year class a(xlO ■•) Stn YA 8-H 8-1- 7-1- 6-1- 5 + 4-1- 3-t- 2-1- \ + Annual Mortality Rate (Mean Shell Length in mm) Stn YA Stn YD 3.7(34.1) 39.2 (35.6) 5.4(31.0) 30.3 (30.8) 5.2 (28.6) 0.0* (26.8) 3.4 (24.7) 16.0 (21.6) 16.7 (23.8) 14.1 (18.9) 21.2(19.2) 34.4 (13.0) 20.0(10.7) 68.1 (7.0) — 86.9 (2.1) 90.0(1.01 87.5 (0.9) Size-Dependent Survivorship of Yoldia notabius 359 TABLE 4. The results of the general linear model for testing the heterogeneity of slopes of survivorship curves hetween stations for each year class and each size class. Factor df F P Year class <1983 3.20 0.0872 1983 1.18 0.2898 1984 0.58 0.4550 1985 0.36 0.5555 1986 0.01 0.9175 1987 0.55 0.4657 1988 15.24 0.0008 1990 0.07 0.7995 Size class (mm) 0-5 0.05 0.8304 10-15 0.85 0.3663 15-20 0.12 0.7371 20-25 0.03 0.8691 25-30 0.58 0.4550 >30 3.02 0.0885 reduced the power of this test. The relationship was highly sig- nificant (/) < 0.001) for the laboratory data. Experiment 1: Size-Specific Crab Predation on the Bivalve Experiment 1 clearly demonstrated that smaller individuals of Y . notabilis suffered higher mortality from crab predation (Fig. 3). Mortality was as high as 80% in the size class 0-5 mm in shell length. It decreased with increasing size and was no more than 20% in the classes larger than 15 mm. Individuals smaller than 10 mm were killed solely by crabs, whereas mortality due to other causes was also observed for those larger than 10 mm. The size- frequency distributions of the bivalve that had been eaten and those that were dead were significantly different from those ex- pected from size-independent predation and mortality, i.e.. the initial size-frequency distribution before the experiment (Kolmo- gorov-Smimov two-sample test; £) = 0.443. p = 0.001 and D = 0.331. p = 0.020. respectively). The burrowing depth of live specimens of Y. notabilis, mea- sured at the end of this experiment, increased with size (Table 6). The results are in agreement with the field and laboratory obser- vations described above (Table 5). The three-factor log-linear model shows a significant interaction only between size and depth (/) = 0.049; Table 7). The nonsignificance in the three-factor interaction among treatment, size, and depth (p = 0.271; Table 7) suggests that the size-burrowing depth relationship was not af- fected by the presence or absence of the predator. Experiment 2: The Effect of Burial Depth on Crab Predation The size-frequency distributions of Y. notabilis in Experiment 2 showed a pattern similar to that in Experiment 1 (Fig. 4). Pre- dation occurred most severely on the smallest size classes (4—10 mm in shell length), even when the burrowing depth was restricted to a depth of 2 cm. The size-frequency distributions of the pre- dated and dead bivalves were not significantly different between the two treatments (Kolmogorov-Smimov two-sample test; D = 0.179. p = 0.971 and£) = 0.133.p = 0.999, respectively). DISCUSSION The annual mortality rate of Y . notabilis in the field was esti- mated to be high (>86%) in the year classes younger than 3 y. In contrast, mortality was lower in most bivalves older than 3 y (Table 3). This leads to the type 3 survivorship curve of Deevey ( 1947). which is typical for marine invertebrates, including several long-lived bivalves such as M. arenaria (Brousseau 1978, Goshima 1982), M. edulis (Thompson 1984). O. edulis (Walne 1961). Mid Scrobicularia plana (Hughes 1970). The only significant between-site difference in the slope of survivorship curves was found in Class 1988 (Table 4), resulting in the large difference in the estimated mortality between stations (Table 3). However, the slopes did not differ between stations when they were compared on the basis of size (Table 4). The relationship between shell length and mortality shows that indi- viduals suffered high mortality when they were smaller than 10 mm in shell length, but this was reduced when shell length ex- ceeded 10 mm (Table 3). The mortality of >'. notabilis thus ap- peared to be more dependent on size than age. Size-dependent mortality has been reported in a variety of ma- rine bivalves, and in most cases, this has been attributed to size- selective predation (Paine 1976, Holland etal. 1980, Arnold 1984, Sanchez-Salazar et al. 1987, Peterson 1990, Seed 1993). In this case, the size-specific mortality pattern in the field corresponded well with that obtained from the predation experiment in the lab- oratory (Figs. 3 and 4). This suggests that predation by the crab, P . liranulata, could contribute to the size-dependent survivorship pattern of Y. notabilis. If one assumes that the feeding rate of P. granulata in the field is similar to that determined in Experiment I (six bivalves mo' ' per crab), this crab would consume about 1 30 bivalves m " "y " ' . This estimate represents 70 and 60% of the natural population at Stn YA and YD, respectively, suggesting that crab predation may have a considerable effect on the survi- vorship pattern of Y. notabilis in the field. Such estimates, however, must be interpreted with caution, especially when one applies the results of laboratory experiments to a field situation. In this study, the relationship between prey density and crab predation rate was not investigated experimen- tally because of difficulties in collecting sufficient numbers of TABLE 5. Depth distribution of Y. notabilis in the field and in the laboratory. No. of Individuals Depth Shell Length (mm) (cm) 0-10 10-20 20-30* 30-40 Field 0-2 4 1 0 1 2-A 1 2 0 1 4-6 0 0 0 4 6-8 0 0 0 0 Laboratory 0-2 11 5 0 0 2-4 0 2 0 7 4-6 0 1 0 3 6-8 0 0 0 1 8-10 0 0 0 0 * Individuals with shell length between 20 and 30 mm did not occur during the study period (January 1990) because of the heterogeneous age structure of the population (Nakaoka 1993). 360 Nakaoka TABLE 6. Depth distribution of Y. notabilis at the end of Experiment 1. (a) Exp #1 : Control o c 0> 3 100-1 80- 60- 40- 20- 0- 10 (b) Exp #1: Experiment 12 0-5 5-10 10-1515-20 20-2525-30 >30 Shell length (mm) Figure 3. Proportion of live and dead individuals of >'. notabilis, main- tained for 28 d in (a) aquaria without predators (Control) and (b) aquaria with the predator P. granulala (Experiment) in Experiment 1. Bivalves were classified into three categories; (1) alive, (2) mortality due to predation (MP; broken shell and unbroken empty shell), and (3) mortality due to other causes (MO; unbroken shell with soft parts remaining). The numeral in each column indicates sample size. crabs and bivalves. Predation rates usually change with prey den- sity, and their relationship (i.e.. functional response) may vary from one environment to another, for example, with sediment type (Lipcius and Hines 1986, Eggleston et al. 1992). The prey den- sities used in this experiment were higher than field densities, and this may change predation rates considerably. Furthermore, the two stations compared in this study differed with respect to the size distribution of Y. notabilis. the composition and abundance of alternative prey, and the sediment composition (Nakaoka 1992b). These differences may shift the prey-preference of foraging activ- ity of the crab, thus influencing feeding rates at the two stations. Predators other than P. granulata may also contribute to the observed survivorship pattern of Y. notabilis in the field. Other potential epibenthic predators include another crab, O. punctatus. and the flounders P. olivaceus and K. bicoloratus, which were collected by dredging and trawling in the study area. However, the No. of Individuals Depth (cm) Shell Length (mm) 0-10 10-20 20-30 30-40 Control 0-2 13 2 0 0 2-A 2 5 3 3 4-6 0 0 4 2 6-8 0 0 1 0 Experiment 0-2 8 4 2 0 2-4 0 6 8 5 4-6 0 0 4 5 6-8 0 0 0 0 densities of these predators could not be estimated by the grab sampler because of their greater mobility compared with P. gran- ulata. Infaunal predators include the naticid snail, Neverita didyma. This snail feeds on Y. notabilis by boring into shell, but its effect is probably minimal, given the low proportion of bored shells in the field (less than 5'7f; Nakaoka unpubl. data). Size-related difference in vulnerability to physical stress or disturbance is another possible factor producing the size- dependent mortality pattern. Levinton and Bambach (1969) esti- mated the size/age-specific mortality of Y. limatula from size dis- tribution of undamaged dead shells and found that individuals smaller than 1 1 mm suffered higher mortality than larger individ- uals in an area with unstable substrate, whereas the mortality was constant over the whole size range in another area with firmer sediments. They considered that the unstable and turbid medium is unfavorable for juveniles, resulting in a size-related change in mortality only in the former site. In Otsuchi Bay, the physical disturbance of bottom sediment is the most severe in the winter, when the northeastern monsoon predominates (Kutsuwada et al. 1988). This may also be related to higher mortality in younger shallow-burrowing individuals. Seasonal changes in survivorship, however, were not obvious because of the variation in monthly estimates of density. In infaunal bivalves, burrowing depth in the sediment is often limited by size, and this is often thought to be responsible for size-dependent mortality (Hughes 1970, Blundon and Kennedy 1982, Goshima 1982). Although the burrowing depth of Y. nota- TABLE 7. The result of the three-factor log-linear model testing the dependence among treatment (with or without predator), shell length, and burrowing depth of live specimens of Y. notabilis at the end of Experiment 1. Source df X^ P Treatment 1 0.53 0.468 Size 3 6.25 0.100 Depth 2 2.98 0.225 Treatment*size 3 3.33 0.343 Trealment*depth 2 1.83 0.401 Size*depth 3 7.86 0.049 Treatment*size*depth I 1.21 0.271 Size-Dependent Survivorship of Yoldia notabius 361 O c 0) 3 (a) Exp #2: 8cm deep (b) Exp #2: 2cm deep 4-10 10-1515-20 20-25 25-30 >30 Shell length (mm) Figure 4. Proportion of live and dead individuals of ) . notabilis, main- tained for 37 d in aquaria with the predator P. granulala and sediment of (a) 8 cm or (b) 2 cm deep in Experiment 2. See legend to Figure 3 for explanation. hilis. an active burrower, was correlated with size (Tables 5 and 6). this relationship was weak when compared with immobile mfaunai bivalves such as M. arenuria (Goshima 1982). Some large individuals of Y. notabilis were found in shallow layers of the sediment, even in the presence of crabs (Table 6). The results of Experiment 1 demonstrate that the size-burrowing depth rela- tionship did not differ significantly between control and experi- mental treatments (Table 7). indicating that Y. notabilis did not change its burrowing depth in the presence of the predator and that crab prcdation was not selectively removing shallow burrowers. Furthermore. Experiment 2 demonstrates that the size-frequency distribution of Y. notabilis eaten by crabs did not differ when the burrowing depth of the bivalve was changed. These findings sug- gest that the size-limited predation is not a consequence of size- dependent burrowing depth in Y. notabilis. The result conflicts with that of similar experiments demonstrating that the predation rate increases when burrowing depth was limited to shallow sed- iments in other bivalves such as M. arenaria (Blundon and Kennedy 1982), M. balthica (Elmgren et al. 1986). and Paphies ventricosa (Haddon et al. 1987). In contrast, Peterson (1990) re- ported that the presence or absence of sediment did not affect the size preference of the crab Callinectes sapidus for the bivalve M. menenaria. In a review of studies of decapod crustacean preda- tion on molluscs, Juanes (1992) found that predatory decapods tend to prefer small-sized prey even when they are capable of feeding on larger prey. Preference for smaller prey is expected when the rates of energy intake decline with size, or when the mechanical or physiological costs of attacking larger prey are too high (Juanes and Hartwick 1990. Juanes 1992). In conclusion. 1 have shown that the survivorship pattern of Y. notabilis is size dependent. In addition, laboratory experiments suggest that size-selective predation by the crab P. granulata may be one of the major factors responsible for the observed size- dependent survivorship pattern. In previous work, 1 have found that spatial variation in food availability controls the variation in the growth rate of Y. notabilis (Nakaoka 1992b). This study sug- gests that variation in growth rates, in turn, leads to variation in survivorship through size-dependent mortality, at least in part driven by crab predation. ACKNOWLEDGMENTS 1 am most grateful to T. Kawamura and K. Morita for their help in the field samplings. 1 thank H. Mukai, S. Ohta, and Y. Shirayama, for their criticisms and discussions throughout the study; E. A. Irlandi, F. Micheli. C. H. Peterson, and A. Tamaki for helpful reviews of the manuscript; and M. Luckenbach for valuable advice on statistical analyses. LITERATURE CITED Arnold. W. S. 1984. The effects of prey size, predator size, and sediment composition on the rate of predation of the blue crab. Callinectes sapidus Rathbun. on the hard clam. Menenaria mercenaria (Linne). J. Exp. Mar. Biol. Ecol. 80:207-219, Bell, S. S. & B. C. Coull. 1980. Experimental evidence for a model of juvenile macrofauna-meiofauna interactions, pp. 179-192. In: K. R. Tenore and B. C. Coull (eds.). Marine benthic dynamics. University of South Carolina Press, Columbia. 451 pp. Blundon, J. A. & V. S. Kennedy. 1982. Refuges for infaunal bivalves from blue crab. Callinectes sapidus (Rathbun), predation in Chesa- peake Bay. J. Exp. Mar. Biol. Ecol. 65:67-81. Brousseau, D. J. 1978. Populalion dynamics of the soft-shell clam Mya arenaria. Mar. Biol. 50:63-71. Deevey, E. S. 1947. Life history tables for natural populations of animals. Q. Rev. Biol. 22:283-314. Eggleston. D. B. R. N. Lipcius & A. H. Mines. 1992. Density-dependent predation by blue crabs upon infaunal clam species with contrasting distribution and abundance patterns. Mar. Ecol. Prog. Ser. 85:55-68. Elmgren, R., S. Ankar, B. Marteleur & G. Ejdung. 1986. Adult interfer- ence wiih postlarvae in soft sediments: the Ponloporeia-Macoma ex- ample. Ecology 67:827-836. Goshima. S. 1982. Population dynamics of the soft clam, Mya arenaria 362 Nakaoka L.. with special reference to its life history pattern. Puhl. Amakusa Mar. Biol. Lab. 6:119-165. Green. R. H. 1973. Growth and mortality in an Arctic intertidal population of Macoma ballhica (Pelecypoda, Tellinldae). J . Fish. Res. Board Can. 30:1345-1348. Haddon, M.. R. G. Wear & H. A. Packer. 1987. Depth and density of burial by the bivalve Paphies veniricosa as refuges from predation by the crab Ovalipes calharus. Mar. Biol. 94:25-30. Holland. A. F.. N. K. Mountford. M. H. Hiegel. K. R. Kaumeyer cS: J. A. Mihursky. 1980. Influence of predation on infaunal abundance in upper Chesapeake Bay, USA. Mar. Biol. 57:221-235. Hughes. R. N. 1970. Population dynamics of the bivalve Scrobiciilaria plana (Da Costal on an intertidal mud-flat in north Wales. J. Anim Ecol. 39:333-356. Jewett, S. C. & H. M. Feder. 1980. Autumn food of adult starry floun- ders. Platichthys slellaius from the northeastern Bering Sea and the south-eastern Chukchi Sea. J. Cons. Int. E.xplor. Mer. 39:7-14. Juanes, F. 1992. Why do decapod crustaceans prefer small-sized mollus- can prey? Mar. Ecol. Prog. Ser. 87:239-249. Juanes. F. & E. B. Hartwick. 1990. Prey size selection in Dungeness crabs: the effect of claw damage. Ecology. 71:744-758. Kennish, M. J. 1980. Shell microgrowth analysis. Mercenaria mercenaria as a type example for research in population dynamics, pp. 225-294. In: D. C. Rhoads and R. A. Lutz (ed ). Skeletal Growth of Aquatic Organisms. Plenum Press. New York. 750 pp. Kutsuwada K.. N. Shikama. T. Kawamura & T. Tanaka. 1988. Results of oceanographic and meteorological observations ( 1988 and 10-year re- ports). Otsuchi Mar. Res. Cent Rep. 14:213-248 (in Japanese). Levinton. J. S. & R. K. Bambach. 1969. Some ecological aspects of bivalve mortality patterns. Am. J. Sci. 268:97-112. Lipcius. B N. & A H. Hines. 1986. Variable functional responses of a manne predator in dissimilar homogeneous microhabitats. Ecology 37: 1361-1371. Moller, P. & R. Rosenberg. 1983. Recruitment, abundance and production of Myij arenaria and Carduim edule in manne shallow waters, western Sweden, Ophelia 22:33-55. Nakaoka. M. 1992a. Age determination and growth analysis based on external shell rings of the protobranch bivalve Yoldia nolabilis Yokoyama in Otsuchi Bay. northeastern Japan. Benthos Res. 43:53- 66. Nakaoka. M. 1992b. Spatial and seasonal variation in growth rate and secondary production of Yoldia notabilis in Otsuchi Bay. Japan, with reference to the influence of food supply from the water column. Mar. Ecol. Prog. Ser. 88:215-223. Nakaoka. M. 1993. Yearly variation in recruitment and its effect on pop- ulation dynamics in Yoldia notabilis (MoUusca: Bivalvia). analyzed using projection matnx model. Res. Pop. Ecol. 35:199-213. Nakaoka. M. & S. Matsui. 1994. Annual vanation in the growth rate of Yoldia notabilis (Bivalvia: Nuculanidae) in Otsuchi Bay. northeastern Japan, analyzed using shell microgrowth patterns. Mcir. Biol. 1 19:397- 404. Paine, R. T. 1976. Size-limited predation: an observational and experi- mental approach with the Mytilus-Pisaster interaction. Ecology 57: 858-873. Peterson, C. H. 1982a. Clam predation by whelks iBusycon spp): exper- imental tests of the importance of prey size, prey density, and seagrass cover. Mar. Biol. 66:159-170. Peterson. C. H. 1982b. The importance of predation and intra- and inter- specific competition in the population ecology of two infaunal suspen- sion-feeding bivalves. Protothaca staminea and Chione undatella. Ecol. Monogr. 52:437-475. Peterson, C. H. 1990. On the role of the ecological experimentation in resource management: managing fisheries through mechanistic under- standing of predator feeding behavior, pp. 821-846. In: R. N. Hughes (ed.). Behavioral mechanisms of food selection. Springer- Verlag. Ber- Im 886 pp. Rhoads, D. C. & D. K. Young. 1970. The influence of deposit-feeding organisms on sediment stability and community structure. J Mar. Res. 28:150-178. Sanchez-Salazar. M. E , C L. Griffiths & R. Seed. 1987. The interactive roles of predation and tidal elevation in structuring populations of the edible cockle, Cerastoderma edule. Estuar. Coast. Shelf Sci. 25:245— 260. SAS Institute. 1987. SAS/STAT® Guide for Personal Computers, Version 6 Edition. SAS Institute Inc., Cary, North Carolina. 1.028 pp. Sasaki. K. 1993. Ecology and stock property of the Sakhalin surf clam. Siii.<:ankenkyii Sosho 42:1-85 (in Japanese with English abstract). Seed, R. 1993. Invertebrate predators and their role in structuring coastal and estuarine populations of filter feeding bivalves, pp. 149-195. In: R. F. Dame (ed). Bivalve filter feeders in esturaine and coastal eco- system processes. Springer- Verlag, Berlin. 579 pp. Seed, R. & R. A. Brown. 1978. Growth as a strategy for survival in two marine bivalves, Cerastoderma edule and Modiolus modiolus. J. Anim. Ecol. 47:283-292. Takagi, Y.. I. Takeuchi, C. Abe, K. Morita, T. Kawamura & A. Tada. 1992. Results of oceanographic and meteorological observations (1992). Otsuchi Mar. Res. Cent. Rep. 18:93-108 (in Japanese). Thompson. R. J. 1984. Production, reproductive effort, reproductive value and reproductive cost in a population of the blue mussel Mytilus edidis from a subarctic environment. Mar. Ecol. Prog. Ser. 16:249- 257. Tyler. A. V. 1972. Food resource division among northern, marine dem- ersal fishes. J. Fish. Res. Board Can. 29:997-1003. Vahl. O. 1981. Energy transformations by the Iceland scallop Chlamys islandica (O. F. Muller), from 70°N. II. The population energy bud- get. J. Exp. Mar. Biol. Ecol. 53:297-303. Walne. P. R. I96I. Observations on the mortality of Ostrea edulis. J. Mar. Biol. Assoc. U.K. 41:113-122. Woodin, S. A. & R. Marinelli. 1991. Biogenic habitat modification in manne sediments: the importance of species composition and activity. Symp. Zool. Soc. Lond. 63:231-250. Zwarts. L. & A. Blomert. 1992a. Why knot Calidris canutus take me- dium-sized Macoma balthica when six prey species are available? Mar. Ecol. Prog. Ser. 83:113-128. Zwarts, L. & A. Blomert. 1992b. Annual and seasonal variation in the food supply harvestable by knot Calidris canutus staging in the Wad- den Sea in later summer. Mar. Ecol. Prog. Ser. 83:129-139. Journal of Shellfish Research. Vol. 15. No. 2, 363-368. 1996. BROWN RING DISEASE AND PARASITES IN CLAMS (RUDITAPES DECUSSATUS AND R. PHILIPPINARUM) FROM SPAIN AND PORTUGAL A. FIGUERAS, J. A. F. ROBLEDO, AND B. NOVOA Instituto de Investigaciones Marinas CSIC. Ediiardo Cabello, 6 36208 Vigo (Spain) ABSTRACT Signs of brown ring disease (BRD) have been detected in Rudiiapes decussatus and Ruditapes philippinamm cultured in Galicia (NW of Spain) and Aveiro (W of Portugal). Clams found on the surface had higher BRD prevalences than did burrowing ones. BRD signs were also detected in R phitippinarum imported from Goro (Italy) into a depuration plant located in the Rfa de Vigo. Histopathological studies revealed the presence of several potential pathogens, the most prevalent being (maximum prevalences given) a chlamydia-like organism (869?). a haplosporidian 1 lOO'/f ). and a PerkinsusAAie organism (60'7f ), Bacterial growth was also found in some of the histologically examined samples. A positive relation between the parasitic load and BRD signs was found. KEY WORDS: Brown nng disease, clams. Rudiiapes decussatus. Ruditapes philippmarum. parasites, acquaculture INTRODUCTION Brown ring disease (BRD) was described as the cause of a mass mortality (more than 80'/f I of cultured clams {Rudiiapes philippi- narunU on the French Atlantic Coast by Paillard et al. ( 1989). The first sign of the disease consisted of the appearance of small brown spots surrounded by a pale brown halo that, in more advanced stages, developed into an abnormal organic brown-black deposit along the pallial line and at the inner edge of the shell. Although Paillard and Maes (1990, 1994) have demonstrated the transmis- sibility of the disease by injecting a bacteria, named Vibrio PI. isolated from diseased clams into healthy clams, there is still some controversy on the etiology of these signs. Some authors have reported that the abnormal deposit of conchiolin in the shell could be a result of the influence of several factors such as contaminants (Alzieu et al. 1981). nutritional deficiencies (Goulletquer et al. 1989). fungi (Alderman & Gareth Jones 1971 ), and parasites (Far- ley 1968. Farley et al. 1988). Juvenile oyster disease, a disease that shares with BRD the appearance of brown deposits on the interior of the shell, has recently been described in hatchery-reared juvenile oysters Crassoslrea virginica (Bncelj et al. 1992) and bacteria were also considered a possible aetiological agent. The production of clams in Galicia (NW Spain) is about 2.0(X) tons per year (2.300 tons in 1991: C.l.P.E.M. 1992) with a rela- tively high economic value ( 1 .500 pesetas per kilogram, I US$ equals 100 Spanish pesetas). Since 1989-1990. repeated mortali- ties have been detected in several clam beds from this area, but their etiology has not yet been established. Previous clam [Rudi- iapes decussatiLS) mortalities in the area have been associated with the presence of Perkinsus- and haplosporidian-like organisms (Figueras et al. 1992). BRD has been reported in Spain on the coast of Cadiz (south of Spain) (Castro et al. 1990. Castro et al. 1992). In this work, we assess the prevalence of BRD in clams taken from several natural beds from northwest Spain and west Portugal. The presence of other potential pathogens in the same sampling locations was established by histology, and the relation between the presence of parasites and BRD signs was also studied. MATERIALS AND METHODS Study Area The epizootiological studies were conducted in natural clam beds in Santa Cristina, Moafia, and Punta Cabalo (Ria de Vigo) and in "storage" clam beds in Carril (Ri'a de Arousa) and Leis (Ri'a de Camarinas) in Spain and Gafanhas (Ria de Aveiro) in Portugal (Fig. I). Samples of carpet-shell clams (/?. deciissaliis) and manila clams {R. philipptnanim) were taken between April and December 1993. According to the location of the clams in the sediment, two groups were distinguishable; (i) burrowing clams (individuals found burtowed in the sand) and (ii) surfacing clams (individuals found on the surface of the sand). A sample of clams, Rudiiapes pullaslra, from Moaiia (Ria de Vigo, Spain) and an- other sample of R. philippinarum imported from Goro (Italy) for marketing in Spain were also examined. .Assessment of the Signs of BRD In each sample location. 100 clams were taken for BRD stud- ies. Clams were opened, cutting the adductor muscles with a scal- pel, trying to avoid any damage to the shell. The inner side of the valves were examined at 25 x with a binocular microscope (Ni- kon) for the appearance of BRD signs. The BRD developmental stage was assessed by the method of Paillard and Maes (1994). This takes into account the spread and thickness of the brown organic deposit in the interior of the shell and the number of affected valves, scoring from stage 1 (BRD signs visible only under a dissection microscope) to stage 7 (BRD signs visible to the naked eye with both valves affected). Histopathological Studies Thirty animals from each sampling location were used for his- tological studies. These animals were fixed whole, after shucking, in Davidson's fixative (Shaw and Battle 1957) for 24 h, and ob- lique transverse sections, approximately 5 mm thick, were taken from each specimen so that mantle, gonad, digestive gland, gills, kidney, and foot tissues were included. Tissue samples were em- bedded in paraffin and, 5-(im sections were stained with haema- toxylin-eosin. The Macchiavello stain for rickettsia was also used (Culling 1974). Clam histological sections were examined for the presence of parasites with the aid of a microscope at 400-1, OOOx magnification (Nikon Optiphot). Statistical Analysis A G-test of independence with contingency tables (Sokal and Rohlf. 1981 ) was used to study the relationship of the presence of 363 364 FiGUERAS ET AL. ■ 42' 30 Atlantic Ocean Ria de Pontevedra i3 I- »!j^^-<--i ^HJ*iinta Cabalo 42- 15' Figure 1. Sampling areas with the particular locations where the epidemiological studies were conducted. Map 1 shows Ria de Camariiias (Galicia, Spain). Map 2 shows Ria de Arousa and Ria de Vigo (Galicia, Spain). Map 3 shows Ria de Aveiro (Portugal). the disease with the sampling site and with clam location (bur- rowing, surfacing) in the sediment. The statistical significance of the relation is expressed as follows: NS, significance (p >0.05); significance. *0.05 < p < 0.01: **0.01 < p < 0.005: ***p < 0.005. To investigate the relation between the different pathogens and the presence of BRD signs, a Pearson correlation matrix with Bonferroni adjusted probability was calculated for each species (Systat). BRD AND Parasites in Clams 365 RESULTS BRD Signs BRD was detected in R. deciissalKs and R philippiinirum cul- tured from all sampled sites in Galicia (Spainl and Gatanhas (Por- tugal). In severely affected clams, the signs consisted of an ab- normally thick deposit of dark brown organic material along the pallial line and at the inner edge of the shell. The percentage of clams affected by BRD varied dependmg on the sampling site, clam species, and clam location (burrowing, surfacing) (Table II. The maximum prevalence of the disease was found in Carril. where almost all examined carpet-shell clams (/?. decussatus) found on the sediment surface were affected. Manila clams (/?. philippinarum) taken from the same area showed considerably lower values, and these differences among clam species were sta- tistically significant (df = I in each case, n = 336, G = 45.55***). In Camarihas. surfacing manila clams were the species most affected by the disease (84.21%). The presence of BRD was de- pendent on the clam species in the burrowing clams (df = I in each case. R. decussatus and R. philippinarum. n = 349. G = 3.97*) and independent in the surfacing ones (df = 1 in each case. R. decussatus and R philippinarum. n = 173. G = 2.85 NS). In Riu de Vigo, the prevalence of the disease signs was quite low in all of the sampled clam beds, with 1 3% being the maximum detected. It is important to point out that BRD signs were also detected in R. pullastra. although with very low prevalence (2.229'f ). In Gafanhas (Portugal), there was a time-related increase in BRD prevalence, mainly in the carpet-shell clams, from March (11%) to November (37%). The most frequent BRD stages detected in all clam species, localities, and situation in the sediment were the second, third, and fourth. Only in the R decussatus found on the surface of the sediment in CamI (prevalence. 93 and 11%) were detected all disease stages (1-7) detected. Histopathological Studies The prevalences of all of the potential pathogens detected are indicated in Table 2. No rickettsiae were observed, despite the use of the Macchiavelo staining method. Chlamydia-like organisms (CLO) consisted of small spherical inclusion bodies (mean diameter. 8.3 |j,m; SD = 1.32; n = 35) found in the cells of the digestive tubules and in the connective tissue. Host cells contain- ing the CLO colonies appeared hypertrophied with a compression of the host cell nucleus against the basal membrane. Although the prevalence was high (up to 86%) and it was found in all sampled TABLE 1. Percentages of BRD in cbms from Spain and Portugal. No. Mean Length Locality Date Clam Species Type of clams (mm) (±SD) BRD(%) Ria de Caniarifias (Galicia. Spain) Leis 6/5/93 R. deciissauis*** S 111 23.0 (±15.7) 62.1 6/5/93 B 106 25.0 (±4.9) 22.6 6/5/93 R philippinarum*** S 19 44.1 (±5.7) 84.2 6/5/93 B 99 35.4 (±8.1) 13.1 16/12/93 R. decictsaliis*** S 13 38.1 (±6.5) 69.2 16/12/93 B 109 38.7 (±3.5) 17.4 16/12/93 R philippinarum*** S 29 43.0 (±5.4) 68.9 16/12/93 B 35 45.8 (±4.1) 8.5 Ria de Arousa (Galicia. Spain) CamI 22/4/93 R decus.<;anis*** S 75 39.1 (±4.0) 93.3 22/4/93 B 98 38.2 (±6.7) 67.3 22/4/93 R. philippinarum B 63 49 (±6.4) 19.1 21/7/93 R decussatus*** S 63 35.5 (±5.4) 77.7 21/7/93 B 99 37.5 (±3.6) 6.7 21/7/93 R. philippinarum B 106 40.7 (±3.6) 6.6 Ria de Vigo (Galicia. Spain) Sta. Cristina 23/4/93 R. decussatus B 103 38.3 (±5.5) 0.9 Sla. Cristina 23/4/93 R. philippinarum Italy 72 44.2 (±3.3) 26.1 Sta. Cristina 24/8/93 R. decussatus B 105 43.0 (±3.4) 12.3 P. Cabalo 1/12/93 R. decussatus B 99 41.7 (±2.4) 9.1 Moana 1/12/93 R. decussatus B 105 45.1 (±4.5) 3.8 Moana 1/12/93 V. pullastra B 45 39.8 (±2.6) 2.2 Ria de Aveiro (Portugal) Gafanhas 3/6/93 R. decussatus NS S 41 37.2 (±2.5) 12.2 3/6/93 B 90 38.0 (±2.5) U.l 3/6/93 R. philippinarum B 90 26.4 (±2.2) 3.3 28/11/93 R. decussatus B 81 40.3 (±5.7) 37.0 28/11/93 R. philippinarum B 101 38.6 (±2.3) 4.9 S. surfacing; B, burrowing; SD, standard deviation; NS and ***, not significant and significant (p < 0.005). respectively, when the type of location (surfacing and burrowing) was compared. 366 FiGUERAS ET AL. TABLE 2. Prevalence of pathogenic agents found in clams from Spain and Portugal. Clam Species Type Pathogens (%) Locality Date CLO BC HAP PER TRE CIL Ri'a de Camarinas (Galicia, Spain) Leis 6/5/93 R. decussatus S 16.6 0 100 0 33.3 10.0 6/5/93 B 20.0 0 96.6 0 40.0 3.3 6/5/93 R. philipptnantm S ND 6/5/93 B 30.0 0 0 0 3.3 0 16/12/93 R. decussatus S* 50.0 0 66.6 0 41.0 0 16/12/93 B 23.3 26.6 83.3 0 56.6 0 16/12/93 R. phiUppinarum S 11.5 53.8 19.2 0 19.2 3.8 16/12/93 B 36.6 30 6.6 0 10.0 0 Ria de Arousa (Galicia, Spain) Caml 22/4/93 R. decussatus S 10.0 0 55.1 6.89 3.4 27.5 22/4/93 B 86.2 0 60.0 60.0 40.0 40.0 22/4/93 R. phiUppinarum B ND 21/7/93 R. decussatus S 66.6 0 16.6 0 0 16.6 21/7/93 B 40.0 0 0 6.6 6.6 3.3 21/7/93 R phiUppinarum B 41.6 0 0 8.3 0 0 Ri de Vigo (Galicia, Spain) Sta. Cnstina 23/4/93 R. decussatus B ND Sta. Cnstina 23/4/93 R. phiUppinarum Italy ND Sta. Cristina 24/8/93 R. decussatus B 33.3 0 0 40.0 3.3 6.6 P. Cabalo 1/12/93 R. decussatus B 26.6 6.6 86.6 23.3 3.3 20.0 Moana 1/12/93 R. decussatus B 13.3 93.3 3.33 10.0 0 0 Moaiia 1/12/93 V. pullastra B 17.4 20.9 0 0 3.4 3.4 Ri'a de Aveiro (Portugal) Gafanhas 3/6/93 R decussatus S 30.0 u 20.(1 0 0 6.6 3/6/93 B 46.6 0 16.6 0 0 3.3 3/6/93 R phiUppinarum B 10.0 0 0 0 0 3.3 28/11/93 R decussatus B 71.4 0 0 0 0 3.5 28/11/93 R. phiUppinarum B 16.6 0 0 n 0 0 BC. bacterial colony; HAP. haplopospondian; PER, Perkinsus-\\ke organism; TRE, tramatode; CIL, ciliate; ND, not done. See Table 1 footnote for other abbreviations. * In the sample of/?, decussatus from Leis (Ria de Camarinas) in December 1993. only 13 clams were analyzed for histological examination. In the rest of the samples, 30 individuals were processed. locations, the intensity was low («5 CLO per section). No host reaction was detected. Bacteiia were found in both R. decussatus and R. phiUppi- narum. This condition was often detected in the gills, connective tissue, and foot muscle. Bacteria were found in pockets scattered throughout tissues. The haplosporidian (mean diameter. 10.75 (j.m. SD = 4.6; n = 30) was found intracellularly in the epithelium of the stomach and intestine, in the cells of the primary digestive tubules, and in the gills. The intensity was very variable. No spores were found. No host response was detected. In Camaritias, R. decussatus showed the highest haplosporidian prevalence, with a year-round value higher than 65%. The clams found on the surface had, in the month of May 1993. a prevalence of 100%. Perkinsus-Wke organisms were found in the connective tissues of the digestive gland and foot and in the gills ofR. decussatus and R phiUppinarum. but the first species always had the higher prev- alences. The mean diameter of the Perkinsus developmental stages detected varied between 3 and 14 |a.m. A strong host response consi.-.Ung of a hemocyte infiltration surrounding the parasite cells was often found. This pathogen was found only in the Ria de Arosa and Ria de Vigo. Trematodes were often found in the foot and in the connective tissue of all of the sampled species in all of the sampled locations, with the exception of the Ri'a de Aveiro (Portugal). These parasites often disrupted the foot muscle, occasionally eliciting a strong host response. Arrested gonadal development was observed in a few individuals. Ciliates were detected at low prevalence in the gill epithelium, and again, no damage or host reaction was observed in association with these organisms. No species identification was attempted. Other potential pathogens detected, with much lower prevalences, were MarteiUa-hke organisms, turbellarians and a gregarine re- sembling Nematopsis sp. In Carril. R. decussatus had the highest prevalence of all of the pathogens, whereas R phiUppinarum showed a very low preva- lence of all of the detected pathogens. In the Ria de Vigo. Perk- insus-Wkt organisms and CLO were the pathogens with the highest prevalence in R. decussatus. R. puUastra was also parasitized, but with lower prevalences than R. decussatus. In the clams from Gafanhas (Portugal), the haplosporidian and CLO also reached high prevalences (20 and 70%, respectively). The sampling locations and the clam species with higher BRD prevalence were also the ones with a higher prevalence of para- BRD AND Parasites in Clams 367 sitism, suggesting a possible relation between both conditions. No statistically significant correlation was found between the presence of BRD signs in the shells of individual clams and the prevalence of the different pathogens found in the histological slides. DISCUSSION BRD is reported for the first time in Camarinas. Vigo (Galicia, NW of Spain), and Gafanhas (Portugal). BRD was not noted in previous studies on the health status of several clam populations in Galicia (Figueras et al. 1992). The BRD signs detected in R. clecussalus and R. philippinarum sampled from several clam beds in Galicia (Spain) and in Gafanhas (Portugal) were identical to those previously described in France and in the south of Spain (Paillard et al. 1989. Castro et al. 1992). The etiology of this disease in European clams has been proved by several authors who were able to reproduce the disease by injecting Vibrio PI into healthy clams (Paillard et al. 1989, Pail- lard and Maes 1990). Although a synergism between the presence of this bacteria in high numbers and other factors such as poor water quality may exist, it is accepted that this Vibrio plays an important role in the development of the disease in the European clam species (Paillard and Maes 1994). Maes and Paillard (1992) studied the effect of Vibrio PI in different species of bivalve molluscs and found that Vibrio PI is more virulent for R. philip- pinarum than for R. decussaius: however. Oubella et al. (1993) showed that Vibrio PI induced an increased density of circulating hemocytes in both clam species. The highest prevalence of BRD was found in Camaririas and Carril. In these areas, clams are kept at a high population density for at least 1 month before shipment to the market. As a conse- quence of these high densities, the food availability is low. In the Ria de Vigo, where the prevalence of BRD was low compared with that in the other sampled sites, the population density was also low. Plana and Le Pennec (1991) demonstrated in laboratory experiments that the mortality attributable to Vibrio is decreased by 40% in fed animals, and they suggested that good nutrition will diminish the detrimental effect of potentially pathogenic bacteria. The frequency of BRD signs was higher in surfacing clams than m burrowing ones in both species. As mentioned by Paillard and Maes (1990), the BRD-affected animals would rise to the surface of the sand before dying, explaining the higher BRD prev- alence in surfacing clams. The distribution of BRD stages in our samples suggests that the evolution of the disease seems to be quite fast in the first steps, with either a subsequent recovery or death of the affected animals thereafter. These results could be explained on the basis of two hypotheses. The first hypothesis stems from the fact that if BRD causes the death of the clam, the animals with advanced disease are likely to die. making it difficult to detect advanced BRD stages in a sample. The second hypothesis is based on the ability of the clam to repair the shell. Paillard and Maes ( 1994) established that after reaching stage 2 of the disease, clams could recover and only the weaker individuals would develop more advanced stages of the disease. As has been observed in France, the temporal differences on BRD prevalences found in this study in several sampled loca- tions suggest a seasonal pattern of the disease (Paillard et al. 1994). Although no correlation was found for individual clams. Ca- marinas, the area where the prevalence of the BRD was highest, also reached the highest prevalences of all detected pathogens, mainly the haplosporidian. The same is true for Carril. but in this case, the Perkinsus-Wke organisms also showed high prevalences. Perkinsiis and haplosporidian species often cause mortalities in cultured molluscs. Perkinsiis marinus and Haplosporidium nelsoni (MSX) are the major diseases of eastern oysters. C. virginica, from the Atlantic Coast of the United States (Andrews 1988. Haskin and Andrews 1988). P. marinus also occurs throughout the Gulf of Mexico. In a previous study on the health status of clams from Galicia, Figueras et al. (1992) associated the presence of a Perkinsus-hke organism with an abnormally high mortality in R. (lecussatus imported from Portugal to a depuration plant in Ria de Vigo. The energetic burden that these pathogens place on their hosts could be used to explain, at least partially, the presence of BRD. because of their weakened situation. Signs similar to those of BRD have been associated with the presence of trematodes in Donax viitariis. Venerupis pullastra. and Ruditapes aureus (Doll- fus 1912. Johannessen 1973. Bartoli 1974). fungus (Alderman and Gareth Jones 1971), and mortalities of unknown etiology (Marin and Dauphin 1991). During this study, trematodes were detected in some clams showing BRD signs. As Goulletquer et al. ( 1989) pomted out. the calcification abnormalities that constitute the ma- jor signs of BRD could result from a "disturbance of protein metabolism, due to mechanical or chemical agents, at the level of amino acid biosynthesis or genetic transcription.'" The high prev- alence of several potential pathogens detected could easily explain this disturbance in the metabolism. Moreover, the isolated strains of Vibrio PI and closely related bacteria (Novoa et al. unpub. obs.) may act synergistically with the parasites (i.e.. depleting the host energy reserves, disturbing several metabolic pathways) to produce the high prevalence of BRD detected in several sampled places. The experimental infections that we are conducting with several Galician Vibrio PI strains and cultured Perkinsus allanti- cus may clarify whether there is any increased susceptibility to BRD in the clams and the events (metabolic, defense mechanisms) that take place in the background when the two pathogens are simultaneously present. ACKNOWLEDGMENTS We thank J. R. Caldas. I. Loureiro. and H. Alvarez for pro- viding technical assistance for histological technics. We acknowl- edge J. Duran. V. Vidal. I. Cunha. M. J. Almeida, and C. Cabin for supplying the clams. J. A. F. Robledo acknowledges the Xunta de Galicia for his fellowship in the IIM-CSIC. This work was financed by the FAR project AQ 3.763 of the EEC. Alderman. D. J. & E. B. Garetti Jones. 1971 Shell disease of oysters. Fish. Invest. Series XXVI(8):19. Alzieu. C. M.. M. Heral. Y. Thibaud. M. J. Dardignac. M. Feuillet. 1981. Influence des peintures antisalissures a base d'organostanniques sur la calcification de la coquille de I'huitre Crassosrrea gigas. Rev. Trav. Insl. Peches Marii. 45(2);101-1 16. LITERATURE CITED Andrews. J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marinus and its effects on the oyster industry, pp. 47-63. In: W. S. Fisher (ed.). Disease Processes in Marine Bivalve Molluscs. Am. Fish. Soc. Spec. Publi. No. 18. Bartoli. P. 1974. Recherches sur les Gymnophallidae F. N. Morozov. 1955 (Digenea). parasites d'oiseaux des coles de Camargue: System- 368 FiGUERAS ET AL. alique. biologie et ecologie. These. University of Aix-Marseille. France. Bncelj. V. M.. S. E. Ford, F. J. Borrero. F. O. Perkins. G. Rivara. R. E. Hillman. R. A. Elston & J. Chang. 1992. Unexplained mortalities of hatchen.' reared, juvenile oysters. Crassoslrea virginica (Gmelin). J. Shellfish Res. 1 1(2);331-347. Castro. D.. M. A. Moriiiigo. E. Martinez. E. Comax & J. J. Borrego. 1990. Microflora associated of "brown ring" from clams [Tapes semi- decussatiis) cultured in southwestern Spain, p. 56. In: A. Figueras (ed.). Abstracts 4th Int. Coll. Pathol. Mar. Aquae. September 1990. Vigo (Spain). Castro. D., E. Martinez-Manzanares. A. Luque. B. Fouz, M, A. Monri- igo. J. J. Borrego & A. E. Toranzo. 1992. Charactenzation of strains related to brown ring disease outbreaks in southwestern Spain. Dis. Aquat. Org. 14:229-236. C. I. P. E. M. 1992. Resultados da Camparia Mariqueira 1991-1992. Secretaria Xeral Tecnica Unidade de Estadi'stica. C. I. P. E M. (ed.l. Xunta de Galicia. Espana. Culling. C. F. A. 1974. Handbook of histopathological and histochemical techniques. 3rd ed. Butterworths. London. Dollfus. R. P. 1912. Contribution a Tetude des trematodes manns des cotes du Boulonnais. Une meta-cercaire margantigene parasite de Donax vittatus Da Costa. Mem. Soc. Zool. Fr 25:85-144. Farley. C. A. 1968. Michinia nelsoni (Haplosponda) disease syndrome in the American oyster, Crassostrea virginica. J. Prowzool. 15:585-599. Farley, C. A., P. H. Wolf & R. A. Elston. 1988. A long term study of microcell disease in oysters with a description of a new genus. Mik- rocvtos (g.n.). and two new species. Mikrocytos mackini (sp.n.) and Mikrocytos roughleyii (sp.n. I. Fish. Bull. 86:581-593. Figueras. A.. J. A. F. Robledo & B. Novoa. 1992. Occurrence of hap- losporidian and Perkinsiis-Vike infections in carpet-shell clams. Riidi- lapes decussatus (Linnaeus. 1758). of the Ria of Vigo (Galicia. NW Spain). J. Shellfish Res. 1 1(2):377-382. GouUetquer. P. H.. M. Heral. C. Bechemin & P. Richard. 1989. Anom- alies de calcification chez la palourde japonaise Rudilapes philippi- nariim: caracterisation et comparaison des compositions en acides amines de differentes parties de la coquille analysees par HPLC. Aqua- culture 81:169-183. Haskins. H. H, & J. D. Andrews. 1988. Uncertainties and speculations about the life cycle of the eastern oyster pathogen Haplosporidium nelsoni (MSX). pp. 5-22. In: W. S. Fisher (ed.). Diseases Processes in Manne Bivalve Mollusc. Am. Fish. Soc. Spec. Puhli. No. 18. Johannessen. O. H. 1973. Deformations of the inner shell surface of Venerupis pullastra (Montagu) (Lamellibranchia) as a result of infec- tion by a trematod metacercaria with a not of parasitism leading to parasitic castration. Sarsia 52:117-122. Maes. P. & C. Paillard. 1992. Effect de Vibrio PI . pathogene de Rudilapes philippinarum sur d'autres especes de bivalves. Haliotis 14:141-148. Marin. F. & Y. Dauphin. 1991. Diversite des alterations dans la compo- sition en acides amines de la phase organique de la nacre des huitres perlieres de Polynesie frangaise (Pinctada margaritifera) atteintes par I'epizootie. C. R. Acad. Sci. Paris 312 Serie 111:483-488. Oubella. R.. P. Maes. C. Paillard & M. Auffret. 1993. Experimentally induced variation in hemocyte density for Rudilapes philippinarum and R decussalus (Mollusca. Bivalvia). Dis. Aquat. Org. 15:193-197. Paillard. C. & P. Maes. 1990. foiologie de la maladie de I'anneau brun chez Tapes philippinarum: pathogenicite dun Vibrio sp. C. R. Acad. Sci. Paris 310. Sene 111:15-20. Paillard. C. & P. Maes. 1994. The brown nng disease symptom in the manila clam. Rudilapes philippinarum: establishment of a classifica- tion system. Dis. Aquat. Org. 19:137-146. Paillard. C. P. Maes. & R. Oubella. 1994. Brown nng disease in clams. Annu. Rev. Fish Dis. 4:219-240. Paillard. C. L. Percelay. M. Le Pennec & D. Picard. 1989. Ongine pathogene de I'anneau brun chez Tapes philippinarum (MoUusque. bivalve). C. R. Acad. Sci. Paris 309. Serie 111:235-241. Plana. S. & M. Le Pennec. 1991. Alterations de la glande digestive et consequences nutritionnelles chez la palourde Rudilapes philippinarum contaminee par une bacterie du genre Vibrio. Aqual. Living Resour. 4:255-264. Shaw, B. L. & H. I. Battle. 1957. The gross microscopic anatomy of the digestive tract of Crassostrea virginica (Gmelin). Can. J. Zool. 35: 325-346. Sokal. R S. & F. J Rohlf 1981. Biometria Blume. Madrid. 832 pp. Joiinuil of Shellfish Raeanh. Vol, 15, No. 2. 369- ."?74. 1996. ISOLATION OF A NATIVE BACTERIAL STRAIN FROM THE SCALLOP ARGOPECTEN PURPURATUS WITH INHIBITORY EFFECTS AGAINST PATHOGENIC VIBRIOS C. RIQUELME,' G. HAYASHIDA,^ R. ARAYA,' A. UCHIDA,^ M. SATOMI,- AND Y. ISHIDA^ ^Deparlamento de Acuicullura Fac. de Recursos del Mar Universidad de Antofagasta P.O. Bax 170 Antofagasta, Chile ^Laboratory Microbiology Depi. of Fisheries Fac. Agriculture Kyoto University Kyoto, Japan ABSTRACT The mass culture of scallops Argopeclen purpuralits (Lamarck. 1819) faces serious problems because of high larval mortalilies. The mam cause of mortality Is the presence of pathogenic bacteria belonging to the genus Vibrio. In this study, the potential inhibitory activity against vibrios was examined for a native bacterial strain identified as Alleromonas haloplanktis. This strain clearly suppressed the growth of Vibrio alginolyliciis and Vibrio aiigiiillunim. two strains that cause severe mortalities in larval cultures of A. purpuraliis. The active inhibitory components were found to be produced dunng the stationary phase of the culture, and they appear to be sensitive to heat. The inhibitory metabolites were precipitated by ammonium sulphate, and they possibly contained a protein- aceous compound. This is the first report of the isolation of Alleromonas species associated with bivalve culture producing antibiotic substances. This strain is potentially useful as a probiotic in the culture of .4. piirpuratiis . KEY WORDS: Alteromoiui.<^ halopUiiikiis, Arf>opeclen purpuraliis. larvae culture, inhibition, probiotic INTRODUCTION In the last few years, efforts have been made to develop aqua- culture technology for the farming of Argopecien purpuratus in Chile. At this time, there are several aquaculturc centers dedicated to the farming of this species along the Chilean coast (Aiken 1993). However, the intensive farming of A. purpuratus in Chile has been hindered by high larval mortalities. These mortalities have been attributed to the presence of pathogenic bacteria (Na- varro et al. 1991). Recently, the species of bacteria thai cause damage to larvae have been identified as Vibrio anguillarum ( VAR) and Vibrio alginolyticus (Pazos et al. 1993. Riquelme et al. 1995b I. Antibiotics have been used for the prevention of the establish- ment of bacterial pathogens and for the treatment of aquaculture species that harbor pathogenic bacteria. The use of antibiotics in aquaculture activities may. however, pose several environmental risks (McPhearson et al. 1991, Spanggard et al. 1993). Research aimed at developing alternative methods for the control of patho- genic bacteria, such as probiotics that are used in terrestrial ani- mals (Conway 1989), is poorly developed for marine organisms. Very few reports on the use of probiotics in aquaculturc have been published (Westerdahl et al. 1991, Austin et al. 1995, Bergh 1995). In addition, dried spray of Tetraselmis suezica has been used as a prophylactic measure (Austin and Day 1990). The de- velopment of such biological control measures is urgently needed because of problems associated with the increased use of antibi- otics in the aquaculture industry (Hansen 1993). A suitable pro- biotic organism should derive from the autochthonous bacteria at the site of application. Riquelme et al. (1994) demonstrated the permanent presence of bacteria in the reproductive organs of A. purpuratus and also re- ported evidence for the vertical transmission of the bacteria. Sev- eral bacteria associated with A. purpuratus broodstock have not shown pathogenic effects on larvae (Riquelme et al. 1995a). sug- gesting that some of these bacteria may be potential beneficial, or probiotic for larvae. In our laboratory, the reproductive organs of several hundred A. purpuratus broodstock have been bacteriologically examined. We detected inhibitory zones produced by bacterial isolates from these molluscs against vibrios such as V. alginolyticus. In this study, we investigate the potential activity of one of these bacterial strains, found to be associated with A. purpuratus in culture, against pathogenic vibrios. MATERIALS AND METHODS Bacterial Isolation Bacteriological analysis of gonads of A. purpuratus broodstock was described by Riquelme et al. (1995b). Briefly, the gonads were extracted and washed externally with 1% benzalkonium chlo- ride. A small incision was made through the surface of the gonads with a heat-sterilized scalpel, and the contents were spread on tryptone soya agar (TSA) (Oxoid) supplemented with NaCl. In a preliminary screening of inhibitory activity against pathogenic vibrios of A. purpuratus larvae, performed following the experi- mental outline provided by Brock et al. ( 1987). one of six strains with a broad inhibitory spectrum was selected. Hereafter, the se- lected strain is referred to as INH. Identification of /A'// Strain The INH strain was identified according to methods presented by Sakata (1989) and Austin (1991). as well as by the use of a 369 370 RlQUELME ET AL. TABLE 1. Source, code, and sensitivity to A. haloplanktis of the different bacteria strains used. Source/Strain Code Sensitivity* Isolation from larval cultures V. alginolylicus A32 S V. alginolylicus A84 S V. anguillarum VAR s A. hydrophila C s Culture collection V. anguillarum 775 s V. anguillarum IFO 13266 s V. alginolylicus ATCC 17749 s Vibrio parahaemolyticus ATCC 17802 s Vibrio damsella ATCC 33539 s Vibrio ordalii ATCC 33504 s Pseudomonas fluorescens IFO 3903 s Shewanella pulrefaciens IFO 3908 s Micrococcus luteus IFO 3333 s Salmonella nphimurium LT-2 s E. coli k-I2 s S. aureus IFO 13276 s Achromobacter sp. IFO 13495 s Morganella morganii ATCC 25830 s Proteus vulgaria IFO 3851 s A. haloplanktis INH R * S, sensitive; R, resistant. miniaturized multitest system API 20E (Analytab). Furthermore, the 16S ribosomal RNA (rRNA) sequence was used. Primers that anneal to the I6S rRNA genes of almost all bacteria were used. Bact27F (5'-AGAGTTTGATCATGGCTCAGA-3'l and Bact530r (5'-GCCAGCAGCCGCGGTAATAC-3') were expected to am- plify a 500-base-pair section (Lane et al. 1985). These oligode- oxynucleotide primers were synthesized with the DNA Synthe- sizer (Gene Assembler Plus; Pharmacia LKB Biotech. Sweden) and purified with the oligonucleotide Purification Cartridge (Ap- plied Biosystem, Inc., Foster City, CA). Polymerase chain reac- tion (PCR) amplifications were carried out in buffer containing 2.0 mM MgCK. 200 mM each dNTPS. 1 .0 mM primers. 25 \i.lm\ Taq DNA polymerase (Kurabo. Japan), and 500 ng/ml of chromosom- al DNA. This mixture was subjected to 25 cycles of I min at 94°C, 2 min at 50°C, and 3 min at 72°C. Chromosomal DNA was pre- pared from cells cultured overnight. Cells were suspended in 200 |jl1 of a lysis solution (10 mM Tris-HCl, 1 nM EDTA. 1% Tnton X-IOO; pH 8). heated for 5 mm at 100°C, and then transferred on ice. After a single chloroform extraction. 5 ^.1 of supernatant was used in a PCR. Amplified DNA was isolated by agarose gel elec- trophoresis and purified with Geneclean II (Funakoshi Co.. To- kyo, Japan). The amplified purified DNA was sequenced with the DNA Sequencer 373A (Applied Biosystems. Inc.) with the Taq DyeDeoxy® Terminator Cycle Sequencing Kit (Applied Biosys- tems, Inc.). Computer analyses of the sequences were performed with DNASlS-Mac-0300 software (Hitachi Software Engineering Co., Yokohama. Japan). Growth of INH Strain in order to determine the optimal range of temperature for growth of the INH strain, the bacterium was inoculated in tubes of tryptone soya broth (TSB) medium (Oxoid) and incubated for 90 h in quadruplicates at 14. 20. 26. and 37°C. Growth was recorded by measunng optical density at 540 nm in a Spectronic 20D spec- trophotometer (Milton Roy Company). Screening of Inhibitory Activity The bactericidal effect of the INH strain was examined by the double-layer method of Dopazo et al. (1988). TSA plates supple- mented with 27c NaCl were spot inoculated with 10 |j,l of over- night cultures of INH strain. After incubation for 24 h at 20°C. the colonies were killed with chloroform vapor (45 min). Thereafter. 100 fxl of a 10- fold dilution of an overnight culture of a tested bacteria in 6 ml of TSB supplemented with 2% NaCl and 0.9% agar was spread over the plates. The tested bacteria included cul- ture collection strains and three strains pathogenic for A. purpu- ratus — V. anguillarum (VAR) (Riquelme et al. 1995a). V. algi- nolylicus. and Aeromonas hydrophila (Riquelme et al. 1996). In- hibition of growth of the tested bacteria (Table I ) around and/or over the macrocolony with a zone of inhibition greater than 5 mm was considered a positive result. INH Cell Treatments Cells of INH harvested from plates of TSA medium were di- vided into three aliquots and subjected to the following treatments. One aliquot was placed in the bottom of a petri dish and exposed to chloroform vapors for 45 min. These were designated dead bacteria; the lack of viability was checked by inoculating the treated bacteria in fresh medium. The second aliquot of the har- vested bacteria was subject to a temperature regime of 60°C for 30 min. and the third aliquot was sonicated for two periods of 10 min in an ultrasonic 50-W series 4710 sonicater (Cole Parmer Instru- ment Co.). Subsequent to these treatments, the cell suspensions of the INH cells were tested for their ability to inhibit the test bacteria by the method described above. Ammonium Sulphate Fractionation Sonicated cells of INH prepared as described above were sub- jected to fractionation with ammonium sulphate following the method described by Scopes (1987). The precipitated materials and supernatant were dialysed overnight in tubings of 12.000 MW. The dialysed materials were tested for bacterial inhibition by the method described above. The inhibition of vibrios in liquid 20 40 60 TIME (h) 100 Figure 1. Growth o{ A. haloplanktis at different incubation tempera- tures. Vertical lines show standard deviation from four replicates. Isolation of a Strain From A. purpuratvs 371 TABLE 2. Growth inhibition ability of different INH ceil treatments against three pathogenic vibrios of marine organisms. Strain/Treatments V. ordalii (ATCC 33504) V. anguillarum (ATCC 775) V. anguillarum (VAR) Live Dead Cells Sonicated Cells Cells T" 60°C Cells + + + + + + + + + + + + + + + + + + + + + + + + + + + . halo larger than 10 mm; + , inhibition. halo smaller than 10 mm; medium (TSB) was also tested. A 50-ml aliquot of the dialysed product of INH was added to test tubes with 5 ml of medium. The inoculum of the test bacteria was 50 (il of the overnight culture (5 X 10'' cells). Growth was recorded by measuring the optical den- sity at 540 nm. The experiment was carried out in quadruplicate. Effect of Supernatant of INH on the Growth of Pathogenic Vibrios To ascertain the extracellular production of inhibitory products, the culture of INH in TSB medium at times of approximately 18, 36, and 66 h, early, middle, and stationary phases respectively, was centrifuged at 3,840x g for 10 min and the supernatant was recovered. The supernatant was filtered through 0.2-|jLm-pore-size filter membranes (Millipore), and the filtrate was added to TSB medium in 100. 50, and 20% v/v dilutions. After inoculation of the test vibrios in the medium, the growth was recorded as de- scribed above. Effect of INH on Larval Survival Experiments were carried out with healthy A. purpuratus lar- vae obtained from a commercial hatchery located in the north of Chile (23°. 25'S-78°38'W). Larvae were not subject to antibiotic treatments. The larvae were conditioned with a stationary-phase inoculum of INH strain (cell densities, ca. 5 x lO*" cells/ml) for 1 and 24 h in a 1-1 volume of sterile seawater filtered through 0.2- (im-pore-size membranes (Millipore). at a density of two to three organisms permilliliter in Duran Schott bottles. After this incuba- tion time, the larvae were netted with a mesh (Nytal; previously treated by 30 min with ultraviolet light), washed several times with sterile seawater, and challenged with the pathogen V . anguilhirum (VAR) from pure culture. This pathogen was isolated from an epizootic of A. purpuratus larval culture (Riquelme et al. 1995a). Unconditioned controls were subject to the same procedure as conditioned larvae but without the addition of the INH strain. The assays were performed in triplicate, according to the method de- scribed by Riquelme et al. (1996). Briefly. A. purpuratus larvae were added (two organisms per milliliter) to sterile seawater fil- tered through 0.2-jjLm-pore-size membranes (Millipore). contained in cell culture plates of 15-ml capacity (30 larvae per treatment) Overnight cultures of the pathogenic strain of V . anguillarum (VAR) in TSB were washed by centrifugation (3,840x g for 15 min) and suspended in marine saline solution. Approximately 10^ and 10* cells/ml were added to separate plates. Unconditioned larvae were also distributed in culture plates as controls. The bio- assays were carried out at 20°C for 24 h. Survival was quantified in an Olympus stereoscopic microscope. Larvae without apparent motion, showing closed valve and velar inactivity, were consid- ered dead. The Tukey test was used to determine the statistically significant effects of treatments (Zar 1984). RESULTS The phenotypical identification of the INH strain revealed that it is Gram-negative and oxidase positive, did not metabolize glu- cose, and required NaCI for growth. This bacterium was tenta- tively identified as Pseudomonas spp. ox Aheromonas spp. and by the technique of 16 sRNA was confirmed as Aheromonas halo- planktis. The results of the growth experiments of the strain INH showed that the strain grew well at 14 and 20°C (Fig. 1). whereas it exhibited poor growth at 26°C and did not grow at 37°C. The screening of the inhibition of several bacteria species re- vealed that the INH strain inhibited the growth of all bacteria tested (Table 1). Autoinhibition was not found. Inhibitory assays with the cell suspensions of strain INH subsequent to different treatments showed that the sonicated pellet of INH maintained the inhibitory activity. The whole-cell pellet contained the inhibitory properties, but the temperature treatment (60°C for 30 min) re- duced the inhibitory effect (Table 2). The growth of V. anguillarum (VAR) and V. alginolylicus was delayed by the components in the first and second fractions of the ammonium sulphate — precipitated INH cells, although a different inhibition pattern of the two strains was found (Fig. 2). The TSB culture supernatant of INH stationary-phase cells delayed the E c o in lU O Z < O c/) CQ < 20 30 TIME (h) Figure 2. Growth of V. anguillarum (VAR) (A) and V. alginolyticus (B) in TSB medium with different fractions of ammonium sulphate treatments. Vertical lines show standard deviation from four repli- cates. 372 RlQUELME ET AL. 0.6 0.5 0.4 0.3 0.2 0.1 0 ^\ - - 0.6 c o 0.5 in v./ 0.4 LU o 0.3 Z < 0.2 CD 0 0.1 0) CD 0 < 0.6 0.5 ^ts^ 100% F 50% F 20% F CONTROL f? 1 > 10 40 20 TIME (h) Figure 3. Growth of V. anguillarum (VAR) in the supernatant of the early (A), middle (B), and stationary (C) growth phases of A. halo- planktis. Vertical lines show standard deviation from four replicates. growth of pathogenic vibrios. The supernatant of INH from early and middle log phase growth stages, however, did not negatively affect the growth of the vibrios (Figs. 3 and 4). The larval survival experiments showed that the precondition- ing of larvae with the INH strain for a short time ( 1 h) was effec- tive for larval protection against pathogenic vibrios (Fig. 5) (P < 0.05). Preincubation for 24 h resulted in no significant differences from the control (P > 0.05). The best result was found with the 1-h bath and with the addition of the pathogen at 10'' cells/ml. in that case, significant differences were observed between treatment and control groups exposed to the pathogens (P < 0.05), but not unexposed controls groups. DISCUSSION The search for and use of beneficial microorganisms to im- prove the production of terrestrial animals are common procedures 1 KStavric et al. 1992. Conway 1989). However, in aquaculture, there are very few studies that attempt to focus on bacteria that prevent the growth of pathogenic organisms (Westerdahl et al. 1991, Olsson et al. 1992, Nogami and Maeda 1992, Bergh 1995. Austin et al. 1995). In recent years, increased interest has focused on the search for suitable probiotics (Douillet and Langdon 1994. Austin et al. 1995). A suitable probiotic organism should derive from the autochthonous bacteria at the site of application. This condition is fulfilled for the INH strain, in that it was isolated from gonads of A. purpiiratus. The INH strain is tentatively identified as A. haloplanktis. It is a psychrophilic and autochthonous marine strain, on the basis of its temperature dependence for growth and the strict requirement of NaCl for growth. It has been reported that Alteromonas species produce antimicrobial substances from marine samples (Barja et al. 1989, Gauthier and Flateau 1976, McCarthy et al. 1985). How- E c o in %_/ UJ U Z < m cc O CO CD < 20 30 40 TIME (h) Figure 4. Growth of V. alginolyticus in the supernatant of the early (A), middle (B), and stationary (C) growth phases of A. haloplanktis. Vertical lines show standard deviation from four replicates. Isolation of a Strain From A. purpuratus 373 100 a: O CONTROL 1 TREATMENTS CONTROL 2 Figure 5. Survival of A. purpuratus larvae challenged with pathogenic V. anguillarum (VAR) in concentrations of 10' (G) and 10' cells/ml (■), after the preconditioning of larvae during 1 h with ,4. haloplanktis (INHl. The controls are (I) unconditioned larvae challenged with the pathogen and (2) unconditioned larvae without the pathogen. ever, there is no information about the presence of Alteromonas species associated with bivalve molluscs producing antibiotic sub- stances against pathogens of marine organisms. The genus Alter- omonas also has been reported to produce other bioactive sub- stances such as metamorphosis inducers (Weiner et al. 1988. Leitz and Wagner 1993). The screening of the inhibition of several bacterial species re- vealed that A. haloplanktis has a broad inhibitory spectrum, in- hibiting Gram-positive as well as Gram-negative bacteria. Autoin- hibition in A. haloplanktis was not found, although this phenom- enon has been reported in other Alteromonas species producing antibiotic substances (Gauthier and Plateau 1976, McCarthy et al. 1985). The inhibition of larval scallop pathogenes V. anguillarum (VAR). V. alginolyticus (strains A32 and A84), and A. hydrophila (strain C) by A. haloplanktis is remarkable because these patho- genic strains are resistant to several antibiotics (Riquelme et al. 1995b. Riquelme et al. 1996). The clinical strains Escherichia coli K12 and Staphylococcus aureus (IFO 13276) were also found to be sensitive to exposure to A. haloplanktis. The active inhibitory compound(s) is produced or excreted by living cells and appear to be contained intracellularly. Furthermore, the inhibitory sub- stanee(s) appears to be proteinaceous in nature, because the activ- ity was lost subsequent to temperature treatment and the active compound(s) was precipitated with ammonium sulphate. While examining the inhibition pattern of sulphate ammonium fractions, it was observed thai probably more than one active inhibitory component exist. The bactericidal components produced by A. haloplanktis are secondary metabolites excreted only in the sta- tionary phase (Figs. 3 and 4). The delayed growth of V. anguil- larum (VAR) and V. alginolyticus in approximately 20 h after the addition of A. haloplanktis supernatant was similar to that ob- served by Olsson et al. (1992). Those authors reported that super- natant from fish mucus bacterial isolates produced a lag period up to 8 h longer than that of the control. Austin et al. (1995) also reported that V. anguillarum was less sensitive to the supernatant of a probiotic strain of V. alginolyticus. as compared with another pathogen (.Vibrio ordalii) that had a rapid decline in the numbers of culturable cells after exposure to the probiotic supernatant. in the experiment of larvae exposure to pathogenic bacteria, it was found that 1 h of preincubation with A. haloplanktis was more protective than 24 h of preincubation. The results indicate that A. haloplanktis was able to protect the larvae from the infection with V. anguillarum (VAR) in a concentration of 10' cells/ml. In a high concentration of 10'' cells/ml, the protection was reduced. In fish, a short lO-min bath with a probiotic bacteria reduced mortality by V. anguillarum from 90 (control) to 74% (Austin et al. 1995). Considering that in the hatchery the pathogens at the beginning of the epizootic are not found in a high concentration, such as 10'' cells/ml. the periodic bath (e.g.. every day) with A. haloplanktis could be a prophylactic measure. The use of A. haloplanktis. or equivalent bacteria, as a means of larval protection against patho- genic bacteria could be promising. However, further research is needed to establish the appropriate conditions, such as cell con- centrations and incubation times, among others, for practical use as a potential probiotic in scallop aquaculture. ACKNOWLEDGMENTS This study was financed by FONDECYT-CHILE Grant No. 92-0997 and partially by Grant 1950-982. The authors acknowl- edge the comments and suggestions on an early version of the manuscript made by Prof. Staffan Kjelleberg. We also thank anon- ymous reviewers for valuable suggestions. LITERATURE CITED Aiken, D. 1993. jCada vez hay mas ostiones! World Aquaculture 24;7-I9. Austin. B. & J. Day. 1990. Inhibition of prawn pathogenic Vibrio spp by a commercial spray-dried preparation of Tetraselmis suezica. Aquacul- ture 90:389-392. Austin, B. (ed.). 1991. Methods in Aquatic Bacteriology. John Wiley & Sons, Chichester. 425 pp. Austin, B., L. F. Stuckey, P. A. W. Robertson, I. Effendi & D. R. W. Griffith. 1995. A probiotic strain of Vibrio alginolyticus effective in reducing diseases caused by Aeromonas satmonicida . Vibrio anguil- larum and Vibrio ordalii. J Fish Dis. 18:93-96. Barja, J., M. Lemos & A. Toranzo. 1989. Purification and Characteriza- tion of an Antibacterial Substance Produced by a Marine Alteromonas Species. Antimicrobial Agents and Chemotherapy, vol. 33. no. 10. pp 1674-1679. Bergh, O. 1995. Bacteria associated with early life stages of halibut. Hippogtossus hipoglossus L., inhibit growth of a pathogenic Vibrio sp. J. Fish Dis. 18:93-96. Brock, T., D. Smith & M. Madigan 1987. Microbiologia. 4th ed. Pren- tice-Hall Hispanomericana, S.A. Mexico. 906 pp. Conway. P. L. 1989. Lactobacilli: fact and fiction, pp. 263-281. In: R. Grubb. T. Midvedt, and E. Norin (eds). The Regulatory and Protec- tive Role of the Normal Microflora. Macmillan Press Ltd.. London. Dopazo. C. P., M. L. Lemos, C. Lodeiros, J. Bolinches, JJ. L. Barja & A. E. Toranzo. 1988. Inhibitory activity of antibiotic producing ma- rine bacteria against fish pathogens. J. Appl. Bacterial. 65:97-101. Douillel, P. & C. Langdon. 1994. Used of a probiotic for the culture of larvae of the Pacific oyster (Crassostrea gigas Thunberg). Aquaculture 119:25^0. Gauthier, M. J. & 0. N. Plateau. 1976. Antibacterial activity of marine violet-pigmemed Alteromonas with special reference lo the production of brominated compounds. Can. J. Microbiol. 22:1612-1619. Hansen. G. 1993. Bacteriology of early life stages of marine fish. Dr. Scient. Thesis. University of Bergen, Bergen, Norway. Lane, D. J., B. Pace, G. J. Olsen, D. A. Sthal. M. L. Sogin & N. R. 374 RlQUELME ET AL. Pace, 1985. Rapid determination of 16S ribosomal RNA sequences for phylogenetic analysis. Proc\ Natl. Acad. Sci. USA 82:6955-6959. Leitz, T. & T. Wagner. 1993. The marine bacterium Aheromonas espe- jiana induces metamorphosis of the hydroid Hydractinia echmala. Mar. Biol. 115:173-178. McCarthy, S. A.. R. M. Johnson, Kakimoto D. & T. Sakata. 1985. Ef- fects of various agents of the pigment (violacein) and antibiotic pro- duction oi Aheromonas luteoviolacea . Bull. Jpn. Soc. Sci. Fisheries 51:1115-1121. McPhearson, R., A. DePaola, R. Ziwno, L. Miles. Jr., M. Motes & A. Guarino. 1991. Antibiotic resistance in gram negative bacteria from cultured catfish and aquaculture ponds. Aquaculture 99:203-211. Navarro, R., L. Sturia, O. Cordero & M. Avendano. 1991. Fisheries and Aquaculture; Chile, pp. 1001-1015. In: Scallops: Biology, Ecology and Aquaculture. Elsevier Science Publisher. Amsterdam. Nogami, K. &. M. Maeda. 1992. Bacteria as biocontrol agents for reanng larvae of the crab Portiinus irituherculatus. Can. J. Fish. Aquat. Sci. 49:2373-2376. Olsson. J. C, A. Westerdahl, P Conway & S. Kjelleberg. 1992. Intes- tinal colonization potential of turbot {Scopluhabnus maximiis) and dab (Limanda /(ma«18 ppt and temperatures «15°C, although they can tolerate temper- ature as high as 35°C and salinity as low as 10 ppt (Mann et al. 1991). Information regarding temperature-salinity tolerance in C. gigas is, however, limited, and the definitive temperature and salinity tolerances of this species have not been established in the laboratory. Therefore, the competence of the Pacific oyster against P. marinus under different salinity and temperature regimes is of particular concern, before its introduction into the mid-Atlantic region. This study evaluates in the laboratory the competence of ♦Corresponding author. triploid and diploid Pacific oysters and eastern oysters against P. marinus under different temperature and salinity conditions. MATERIALS AND METHODS Experiment I: Temperature Effect Eastern oysters, C. Virginia (shell length [SH], 7-8 cm), were collected on January 8, 1992, from Ross Rock in the Rappanhan- nock River, a tributary of the lower Chesapeake Bay. Oysters from this area typically have a low prevalence of P. marinus infection (Burreson 1992, Ragone Calvo and Bun-eson 1994, Ragone Calvo and Burreson 1995). The ambient temperature and salinity at the time of collection were 8°C and 10 ppt. Triploid (3N, assayed to be 957f| and diploid (2N) Pacific oysters (age, 16 mo; SH, 6-7 cm) were progenies from a spawning conducted by Dr. Standish Allen (Haskin Shellfish Laboratory, Rutger's University) in late July of 1990. The spawning was produced from second-generation parents of 1989 broodstocks from Washington, and juveniles were raised at the Virginia Institute of Marine Science, in quarantined flumes with flowing raw York River water (YRW ambient tem- perature, 8°C and salinity, = 20 ppt, at the time of expenment). Before the start of the experiment, initial assessment was per- formed on a subsample of 20 C. gigas and 25 C. virginica for P. marinus infection by use of the tissue thioglycollate assay (Ray 1952. Ray 1966) described below. All groups tested negative. The remaining C. virginica and C. gigas were held separately in aer- ated 55-gallon tanks and gradually adjusted to the three test tem- peratures ( 10. 15, and 25°C. 2°C per 2 d) at salinities of 20-22 ppt (1 |xm filtered YRW). Before temperature adjustment, C. virgin- ica was first acclimated (3 ppt per 2 d) from ambient salinity (i.e., 10 ppt) to the experimental salinity (i.e., 20-22 ppt). After ad- justment to the desired test temperatures and YRW salinity, oys- ters were maintained in aerated 1 p-m filtered YRW in 40-1 375 376 Chu et al. aquaria (20-22 oysters per aquarium). Oysters were fed with algal paste (a mixture of Tahitian Isochnsis galbana and Thalassiosira pseudonaiia, 0.1 g/oyster) daily, and mortality was recorded throughout the course of the experiment. If oysters died at the beginning of temperature adjustment, they were replaced. Thus, the number of oysters among groups was similar (N = 37^1) when P. marinus challenge was initiated. All experimental oysters were challenged twice with freshly isolated P. marinus meronts. Twenty-nine days after the initiation of temperature acclimation, oysters were inoculated with 0.1 ml of meront/merozoite suspen- sion (2.5 X 10*^ meronts/oyster) into the shell cavity. Control oysters were inoculated with filtered YRW (0.22-fxm-pore-size filter). Forty-one days after the first challenge, challenged oysters were inoculated with a second dose of meronts (7.0 x 10^ meronts per oyster). Sixty-eight days after the first challenge (27 d after the second challenge), 10 control and 10 challenged oysters from each temperature treatment were sacrificed and rectal tissues were re- moved to determine infection by use of the tissue thioglycoUate assay (Ray 1952, Ray 1966). Eighty-four days after the first chal- lenge, the remaining oysters were sacrificed and the same param- eters mentioned above were measured. Data from the two sam- plings were pooled to determine the disease prevalence and inten- sity. Experiment 2: Salinity Effect The experimental protocol of this experiment was similar to that of the temperature effect experiment. C. virginica (7-8 cm) were collected on May 1 1 , 1992. from Ross Rock, Rappahannock River (Ambient temperature, 19°C; salinity. 6 ppt). C. gigus (3N and 2N, 6-8 cm) was from the same stock used for the temperature effect experiment. Initial assessment of P. marinus infection on 20 C. gigas and 25 C. virginica showed that, with the exception of a single P. marinus cell detected in one of the diploid C. gigas. no oysters were infected with P. marinus. The ambient temperature and salinity of YRW at the time of the experiment were 19-22°C and 20 ppt respectively. Both C. virginica and C. gigas were placed in aerated 200-1 tanks, and salinities were gradually ad- justed (3 ppt per 2 d) to salinities of 3, 10, and 20 ppt. at 19-22°C. After salinity adjustment was completed, oysters were maintained in aerated 40-1 aquaria. During the salinity adjustment period, heavy mortality occurred in both diploid and triploid C. gigas at 3 ppt. Consequently, the susceptibility of C. gigas and C. virginica to P. marinus was compared only at 10 and 20 ppt. As in exper- iment 1, test oysters were challenged twice by freshly isolated meronts/merozoites (2.0 x 10^^ cells/oyster, 21 d after the initia- tion of salinity adjustment and 5.0 x 10' cells per oyster 1 2 d after the first challenge). Again, control oysters were inoculated with filtered YRW. Fifty days after the initial P. marinus challenge, the experiment was terminated to determine disease prevalence and intensity. Preparation of MerontlMerozoite Suspension Fresh meront/merozoite suspension was prepared according to La Peyre and Chu (1994). Briefly. P. mannwi-infected oyster tissues were rinsed thoroughly with filtered (0.22 |jLm) YRW and subsequently homogenized in (0.22 |j.m) filtered YRW with a blender (Virtis. Model 23) at high speed for 2 min. The suspension was then passed through a series of screens ( 100, 35. 20, and 15 |j,m) to remove oyster tissue residues. The number of merozoites in suspension was counted with a hemacytometer and adjusted to the desired concentration. P. marinus Assay The tissue thioglycoUate assay (Ray 1952, Ray 1966) was used for P. marinus diagnosis. Rectal tissue was removed from each oyster and incubated in fluid thioglycoUate medium for 4-5 d. The intensity of infection was ranked as 0 (negative). 1 (light). 3 (mod- erate), and 5 (heavy), on the basis of the number of stained P. marinus hypnospores contained in the oyster rectal tissue smear. Statistical Analysis Logistic regression and log-linear modelling (Agresti 1990) were used to determine differences in infection prevalence be- tween temperature and salinity treatments and between oyster spe- cies. Two-factor analysis of variance was used to determine dif- ferences in infection intensity between the three groups (i.e., C. virginica. C. gigas 2N and 3N) of oysters at different temperature or salinity treatments. RESULTS Experiment 1 Mortality Throughout the course of the experiment, a total of 18 C. virginica. 38 diploid (2N) C. gigas. and 39 triploid (3N) C. gigas died. Most of the deaths occurred at 25°C during temperature adjustment (32 triploid C. gigas. 1 diploid C. gigas, and 4 C. virginica) (Table 1). High mortality was also noted at 25°C after oysters were challenged with freshly isolated P. marinus, with the TABLE 1. Mortality of C. virginica and C. gigas During Temperature Acclimation and After Challenge with P. marinus (Dermo). C. virginica C gigas (2N) C. gigas (3N) Mortality 10°C (N = 80) LS°C (N = 80) 25°C (N = 80) ICC (N = 79) LS'C (N = 81) 25°C (N = 85) 10°C (N = 82) 15°C (N = 82) 25°C (N = 111) Mortality (no. of deaths) during acclimation Mortality (no. of deaths) after P. marinus exposure Total mortality (%) during experiment* 0 1 1.3 0 3 3.8 4 10 17.5 0 1 1.3 2 7 11,1 7 21 32.9 1 1 2.4 1 3 4.9 32 1 29,7 no. of dead oysters/initial total number of oysters. Disease Susceptibility of Two Oyster Species 377 exception of'triploid C. fiigas (heavy mortality occurred only at the time of temperature adjustment). Although 21 diploid C. gigcis and 10 C. virginica died, only one triploid C. gigas died at that tem- perature. Unfortunately, no tissue was able to be recovered from some of these mortalities for P. marinus diagnosis. Hence, mor- talities with no meat recovered were excluded from prevalence and intensity calculations. However, for those mortalities that had tis- sues, it was found that one P. wurmwi-challenged C. virginica (N = 9) at 25 °C, one control diploid C. gigas (N = 4) at 15T. and one control (N = 8) and three challenged diploid C. gigas (N = 7) at 25°C were infected. None of the triploid C. gigas (N = 2) that were examined had infections. Prevalence and Intensity of P. marinus Infection Infection prevalence (percentage of infected oysters = number of infected oysters/total number of oysters at the time of inocula- tion) significantly increased (p = 0.(3001) with temperature in all P. /7Mn>iH.$-challenged oysters (Fig. 1). Prevalence was higher in C. virginica than m the two C. gigas groups, with the exception of the 10°C treatment. At 10°C, 3N C. gigas had a higher prevalence (309?-) than both 2N C. gigas (24%) and C. virginica (25%). The infection prevalences at 15 and 25°C. respectively, were 50 and 60% for C. virginica. 36 and 51% for 2N C. gigas. and 37 and 56% for 3N C. gigas. However, these differences were not sta- tistically different (p > 0.05). Infection intensity increased signif- icantly with increase in temperature and was significantly higher (p = 0.001) in C. virginica than C. gigas (2N and 3N) (Fig 2A). At 25°C, 10 (277f) of the infected C. virginica had moderate infections and 5 (14%) had heavy infections. There were four (11%) infected 2N C. gigas at 25°C and one (3%) at lOT with moderate infections. None of the infected 3N C. gigas developed advanced (i.e.. moderate or heavy) infections. Infection intensity expressed as weighted prevalence ( = sum of disease code num- bers/number of oysters) also significantly increased with increas- ing temperature (p = 0.0001). C. virginica had significantly 10 16 26 10 15 26 10 16 26 Temperalura ( C) C. virginica C gigas (2N| C gigas (3N) 41 39 39 10 15 25 10 IS 25 T*mp«ratur« ( C) 10 15 26 Figure 2. Intensity of P. marinus infection in P. morinus-challenged ( Al and control (B) C. virginica and f . gigas (2N and 3N) oysters at 10, 15, and 25°C. Numbers above the bars represent total number of ovsters in each treatment. 10 15 25 10 15 25 Temperature ( C) 10 ^1 Control (N -35-41) If/A Chillcnged (23-41) Figure 1. Prevalence of P. marinus infection (% infected oysters) in control and P. marmus-challenged C. virginica and C. gigas (2N and 3N) ovsters at 10, 15, and 25°C. 378 Chu et al. higher weighted prevalence (p = 0.0004) than C. gigas (2N) and C. gigas (3N). Mean weighted prevalences were 0.79, 0.45, and 0.41 in C. virginica. C. gigas (2N), and C. gigas (3N), respec- tively. Weighted prevalence in 2N and 3N C. gigas was not sta- tistically different. Some of the oysters in the control groups of C. virginica and 2N C. gigas were infected with P. marinus (Fig. 1). Among 2N C. gigas. one oyster (3%) at 15°C and five oysters (13%) at 25°C had light infections. Among C. virginica. nine (24%), four ( 1 1%), and three (8%) oysters had light, moderate, and heavy infections, re- spectively (Fig. 2B). None of the control 3N C. gigas oysters were infected. Experiment 2 Mortality During salinity adjustment, high mortality occurred in both diploid (2N) and triploid (3N) C. gigas. especially when salinity was adjusted down to 3 ppt (44 of 80 diploid died. 37 of 80 triploid died), but no mortality was noted in the C. virginica groups (Table 2). As a result, the C. gigas (2N and 3N) at 3 ppt treatments were terminated. When the dead oysters were examined for P. marinus infection, one 2N and one 3N C. gigas had light infections. After P. marinus challenge, mortality in Pacific oysters was consistently high. In total, 2 challenged and 5 control 2N C. gigas at 20 ppt, 9 challenged and 1 1 control 2N C. gigas at 10 ppt. 9 challenged and 8 control 3N C. gigas at 10 ppt, and 14 control and 3 chal- lenged 3N C. gigas oysters at 20 ppt perished. However, only two control and two challenged eastern oysters died after P. marinus challenge. None of these dead oysters were found to be infected by P. marinus. Prevalence and Intensity of P. marinus Infection In all salinity treatments, C. virginica had the highest preva- lence of P. marinus infection (p = 0.001) (Fig. 3). In the P. /?ifln;!(«-challenged oysters, the prevalence in 2N C. gigas, 3N C. gigas, and C. virginica, respectively, were 25, 35, and 65%' at 10 ppt and 25, 31, and 64% at 20 ppt (Fig. 3). Among the control oysters, no Pacific oysters at 10 ppt were infected, but at the same salinity, 7% of the C. virginica were infected. At 20 ppt, 5% of the 2N C. gigas controls and 13% of the C. virginica controls were infected, whereas none of the 3N C. gigas were infected. Preva- lence was low in C. virginica at 3 ppt, 7 and 3%, in challenged and control groups, respectively. All infected oysters in all groups had only light infections, with the exception of one eastern oyster at 20 ppt, which was moderately infected (Fig. 4). Similar to the results 10 20 Salinity (ppt) ^B Control (N = 2e-40) p7~l Cliill»ng»J (23-41) Figure 3. Prevalence of P. marinus infection {% infected oysters) in control and P. marinus-challenged C. virginica and C. gigas (2N and 3N) oysters at 3, 10, and 20 psu. in the temperature experiment, C. virginica had significantly higher weighted prevalence than C. gigas (2N and 3N) (p = 0.0001). Mean weighted prevalences for C. virginica. C. gigas (2N), and C. gigas (3N) were 0.64, 0.25, and 0.33, respectively. Salinity ( 10 and 20 ppt) did not significantly affect (p > 0.05) the weighted prevalence. In both oyster species, no differences were observed in pooled infection intensity between salinities (10 and 20 ppt). DISCUSSION The results of this study revealed that C. gigas, both diploid and triploid, is less susceptible to P. marinus than is C. virginica. This is consistent with previous findings in experiments comparing P. marinus susceptibility, mortality, and growth rates between C. virginica and C. gigas challenged with the parasite (Meyers et al. 1991, Barber and Mann 1994). At all tested temperature-salinity regimes, P. marinus-chdWengcd C. virginica suffered higher in- fection rates than C. gigas. Although 27 and 14% off. marinus- challenged C. virginica advanced, respectively, to moderate and heavy infections, only 3-11% of moderate infections were de- tected in P. mfln«(/.s-challenged diploid C. gigas. However, be- cause much higher infection rates were found in the control, non- P. marwHi-challenged C. virginica. at any given temperature and salinity treatment, than in non-P. man/i(«-challenged diploid and triploid C. gigas. the authors believe that part of the recorded TABLE 2. Mortality of C. virginica and C. gigas During Salinity Acclimation and After Challenge with P. marinus (Dermo). C. virginica C. gigas (2N) C. gigas (3N) 3 psu 10 psu 20 psu 3 psu 10 psu 20 psu 3 psu 10 psu 20 psu Mortality (N = 79) (N = 84) (N = 81) (N = 77) (N = 86) (N = 95) (N = 52) (N = 78) (N = 81) Mortality (no. of deaths) during acclimation 0 0 0 44 6 9 37 12 10 Mortality (no. of deaths) after P marinus exposure 4 0 0 — 20 7 — 17 17 Total monality (%) dur-ng i;xperiment* 5 0 0 — 30.2 16.8 — 37.1 33.3 No. of dead oysters/initial total number of oysters, — = treatments were terminated before P marinus exposure due to heavy mortalities. Disease Susceptibility of Two Oyster Species 379 £iiy HEAVY 10 20 Sallnily (ppl) / C gigos (2N) C gigas (3N) 39 MODERATE 10 20 10 20 Salinity (ppt) Figure 4. Intensity of P. marinus infection in P. marinus-challenged (A) control (B) C virginica and C. gigas (2N and 3N) oysters at 3, 10, and 20 psu. Numbers above the bars represent total number of oysters in each treatment. infection in P. n!(7/-!/!/(.v-chaIlenged C. virginica was attributed to the expression of hidden infection carried over from the field. Unfortunately, the thioglycollate tissue assay used in this study for P. marinus diagnosis was not sensitive enough to detect cryptic infections, thus restricting the interpretation of the experimental results. However, the nonchallenged C. virginica showed substan- tially lower P. marinus infection prevalence and intensity than did P. mannui-challenged C. virginica. The observed increased dis- ease prevalence and intensity in the challenged C. virginica must have been derived from the laboratory challenge. Future studies should use eastern oysters from an area free of P. marinus for this kind of study. Also, in oysters collected in winter months, over- wintering infections will not develop to detectable levels until 1-2 mo post-exposure to high temperatures (i.e., 25°C). Therefore, to establish baseline information, it would be wise to expose oysters collected during winter to warm temperatures for 1-2 mo before the initial infection assessment. Although Pacific oysters appear less susceptible than eastern oysters to P. marinus infection, heavy non-f. marinus-Te\dled mortality occurred in both diploid and triploid Pacific oysters at salinities of 10 ppt and below and temperatures higher than 15°C during the acclimation period. This indicates that the Pacific oyster may be less tolerant to high-temperature and low-salinity exposure than eastern oysters. High non-disease-related mortality (70%) was also recorded in Pacific oysters, in conjunction with salinities below 20 ppt. in a study carried out to compare the growth and mortality of C. gigas and C. virginica challenged with P. marinus (Barber and Mann 1994). It seems that low salinity exerts a greater effect on the physiology of this species than does high tempera- ture. All C. gigas died when salinity was reduced to 3 ppt. These results suggest that salinities lower than 20 ppt stress C. gigas, thus reducing its resistance to P. marinus. In conclusion, the Pacific oyster. C. gigas. is less susceptible to P. marinus than is the eastern oyster. C. virginica. However, they may not survive if introduced into Chesapeake Bay tributaries because they are unable to adapt well to the low-salinity and high- temperature conditions. The mid-Atlantic climate is relatively warm, between temperate and subtropical. The ecosystem of the Chesapeake Bay is complex. The salinity range of oyster habitats in the Chesapeake Bay varies seasonally, from as low as 0 to >20 ppt (Andrews 1988. Ragone Calvo and Burreson 1995). The water temperature of most tributaries along the bay can reach 28-29°C (Andrews 1988) and persist for more than 2 mo during the sum- mer. The oyster pathogen, P. marinus. on the other hand, can survive in salinities lower than 5 ppt, and epizootics caused by this parasite increase at high temperatures (Andrews 1988, Burreson and Andrews 1988). Moreover, the shells of C. gigas held in water from the lower Chesapeake Bay (i.e.. York River. VA) were found to be quite susceptible to invasion by the polychaete. Poly- dora sp. (Burreson and Mann 1994). Further studies are needed to ascertain the competence of C. gigas to support a commercial fishery in Chesapeake Bay. ACKNOWLEDGMENT This study was made possible in part through funding from the Oyster Disease Research Program. National Marine Fisheries Ser- vice. NOAA grant (# NA90AA-D-FM739). The authors thank Dr. Stanish Allen. Haskin Shellfish Laboratory. Rutger's Univer- sity, for the contribution of 3N and 2N C . gigas juveniles; Drs. Bruce Barber and Roger Mann and their associates for raising and maintaining the C. gigas: and Mr. Kenneth Walker for collection of C. virginica. The invaluable help of Dr. Robert Diaz in statis- tical analyses. Drs. Robert Hale and Kenneth Webb in editing the first draft of the manuscript, and the helpful comments from Drs. Hale. Webb, Barber, and the anonymous reviewers are greatly appreciated. Contribution number 1992 of the Virginia Institute of Marine Science. LITERATURE CITED Agresti, A. 1990. Categorical Data Analysis. John Wiley & Sons. New York. pp. 79-129. Andrews. J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marinus and its effects on the oyster industry . Am. Fish. Soc. Spec. Puhl. 18:47-63. Barber. B. J. & R. Mann. 1994. Growth and mortality of eastern oysters, Crassostrea virginica (Gmelin 1791), and Pacific oysters, Crassostrea gigas (Thunberg. 1793) under challenge from the parasite. Perkinsus marinus. J. Shellfish Res. 13:109-114. Burreson. E. M. 1992. Status of the major oyster diseases in Virginia — 1 99 1 . A summary of the annual monitoring program . Marine Resource Report 92-1. Virginia Institute of Marine Science. Gloucester Point. Virginia. Burreson. E. M. & J. D. Andrews. 1988. Unusual intensification of Ches- 380 Chu et al. apeake Bay oyster diseases during recent drought conditions. Proc. Oceans 88:799-802. Burreson. E. M. & R. Mann. 1994. Field exposure of tnploid Crassoslreu gigas to Haplosporidium nelson! (MSXl and Perkinsus mariniis (Dermo) in tlie lower Chesapeake Bay. J. Shellfish Res. 13:293. Elston, R., M. Kent, & M. Wilkinson. 1987. Resistance of Ostrea ediilis to Bonamia ostrea infection. Aquaculture 64:237-242. Grizel, H. 1985. Etude des recentes epizootics de I'huitre plate Ostrea ediilis L. et de leur impact sur I'ostreiculture bretonne. These de Doc- torat. Universite des Sciences Techniques du Languedoc. Montpellier. France . Gnzel, H., E. Mialhe. D. Chagot.. V. Boula & E. Bachere. 1988. Bon- amiasis: A model study of diseases in marine molluscs. Am. Fish. Soc. Spec. Publ. 18:1-14. La Peyre, J. F. & F. L. E. Chu. 1994. A simple procedure for the isola- tion of Perkinsus marinus merozoites. a pathogen of the eastern oyster, Crassostrea virginica. Bull. Eur. Assoc. Fish. Pathol. 14:101-103. Mann, R., E. M. Burreson & P K Baker. 1991. The decline of the Virginia oyster fishery in Chesapeake Bay: considerations in the intro- duction of a non-endemic species, Crassostrea gigas (Thunberg 1 7931. J. Shellfish Res. 10:379-388. Myers, J. A., E. M. Burreson, B. J. Barber & R. Mann 1991. Suscepti- bility of diploid and triploid Pacific oysters, Crassostrea gigas. to Perkinsus marinus. J. Shellfish Res. 10:433-437. Ragone Calvo, L. M. & E. M. Burreson. 1994. Characterization of over- wintering infections of Perkinsus marinus (Apicomplexa) in Chesa- peake Bay oysters. J. Shellfish Res. 13:123-130. Ragone Calvo, L. M. & E. M. Burreson. 1995. Status of the major oyster diseases in Virginia — 1995. A summary of the annual monitoring pro- gram. Marine Resource Report 96-1 , Virginia Institute of Marine Sci- ence, Gloucester Point, Virginia. Ray, S. M. 1952. A culture technique for the diagnosis of infections with Dermocystidium murinum Mackin. Owen and Collier in oysters. Sci- ence 116:360-361. Ray, S. M. 1966. A review of the culture method tor detecting Dermo- cystidium mariiuim. with suggested modifications and precautions. Proc. Natl. Shellfish. Assoc. 54:55-66. Journal of Shfllfish Research. Vol. 15, No. 2, 381-384. 19%, SPATIAL DISTRIBUTION AND INTENSITY OF PERKINSUS MARINUS INFECTIONS IN OYSTER RECOVERY AREAS IN MARYLAND GUSTAVO W. CALVO, ROBBIE J. PAGAN, KELLY N. GREENHAWK, GARY F. SMITH, AND STEPHEN J. JORDAN Maryland Department of Natural Resources Cooperative Oxford Laboratory 904 S. Morris St. Oxford. Maryland 21654 ABSTRACT The project described here reports on the status and spatial variability of Perkinsiis manim.s in three oyster recovery areas (ORAs) during 1994. The objectives of the study were to e.\amine spatial variability in P. marinus infections among oyster bars along single tributaries and within single oyster bars. In addition, the study was conducted to compare estimates of prevalence based on intensive and spatially accurate patenl-tong sampling with estimates of prevalence based on traditional dredge sampling of 30 oysters. For comparative purposes, patent-tong and dredge sampling were both conducted in each of four oyster bars, within 35 days, during September-October 1994. In June 1994. a pilot study was conducted to test the use of long range navigation (LORAN) and global positioning system (GPS) for the accurate deployment of patent-tongs and to compare hemolymph and tissue Ray tluid thioglycollate medium assays. In summary, results from this investigation support three conclusions: ( I ) During fall 1994, the variation off. marinus prevalence, among and within oyster bars in sampled ORA tributaries, was small l<30'7r) but statistically significant; (2) traditional dredge samples of 30 oysters per bar provided estimates of prevalence remarkably similar (within 5%) to the ones obtained from patent-tong samples of an average of 340 oysters per bar; (3) RFTM assays conducted in the spring showed that hemolymph, rectum, and combined gill and palp samples gave equivalent determinations of prevalence. KEY WORDS: Spatial distribution. P marinus. variability, oysters, management, Maryland INTRODUCTION Perkinsu.\ inanniis (Mackin, Owen, and Collier) is a conta- gious pathogen of oysters Crassostrea virginica (Gmelin). The transmission of the parasite occurs directly, from oyster to oyster, through the surrounding water. Studies by Mackin (1962) and Andrews ( 1988) have indicated that heavy parasite loads in dying oysters spread infections to adjacent oysters. Epizootics are be- lieved to be favored by continuous and densely populated oyster bars in relatively high-temperature and high-salinity environments (Mackin 1962. Ray 1987). According to Andrews 11988). the parasite thrived in beds with high oyster density, and it persisted in public areas with low oyster density. Apparently, disease-induced mortality was highest in areas with high oyster density, relatively isolated areas with low oyster density restricted the spread of the parasite. The spread of this pathogen has also occurred by the movement of infected oysters (Andrews 1988. Ford 1992). Despite extensive studies on the epizootiology of P. marinus reviewed by Andrews ( 1988). there have been few reports on the geographical distribution and variability of infections in oysters within and between oyster bars. Craig et al. (1989) reported re- gional foci of infections in the Gulf of Mexico within scales of 300 km or less and 1 ,500 km or more, and uneven (patchy) distribution of infections with uninfected oysters immediately adjacent to in- fected ones. In a study of the spatial distribution of P . marinus in two oyster reefs in Texas, however. White et al. (1989) reported that infected oysters were evenly distributed on both reefs, even when oysters were unevenly distributed (in patches) in one of the reefs. Management strategies that could possibly prevent or reduce the effect of diseases on oysters require seasonal and geographical monitoring of disease agents and associated pathologies in oysters. Andrews and Ray ( 1988) delineated management strategies to con- trol P . nuinnus in the Chesapeake Bay and the Gulf of Mexico. Proposed strategies included isolating protected oyster beds from surrounding infected areas, harvesting early and removing all oys- ters at harvest, prohibiting the movement of infected oysters, and transferring only disease-free oysters into low-salinity areas, where the acquisition and progression of infections is minimal. In Maryland, oyster recovery areas (ORAs) have been designed to incorporate some of the P. marinus-e\c\ui'ion strategies pro- posed by Andrews and Ray (1988). ORAs are part of a plan to balance environmental and commercial interests in oyster rehabil- itation (Maryland Oyster Roundtable Action Plan 1993). ORAs have been designated in various tributaries of the Chesapeake Bay including the Chester. Choptank. and Nanticoke Rivers (Fig. 1). Those rivers have been divided into two to three zones along the salinity gradient. Zones A correspond to upstream areas with the lowest salinity suitable for oyster habitat. Oyster harvest and the movement of infected oysters into zones A are prohibited. Zones B are located immediately downstream from zones A. Oyster har- vest is permitted within zones B. but the introduction of infected oysters into zones B is not allowed. Zones C. immediately down- stream from zones B. extend to the mouth of the tributaries. At this time, there is no restriction on harvest or on the movement of oysters into zones C. However, if quarantine restrictions in zones B prove effective, similar restrictions may be applied to zones C (Maryland Oyster Roundtable Action Plan 1993). The project de- scribed here reports on the status and spatial variability of P. marinus in three ORAs during 1994. The objectives of this study were to examine spatial variability in P. marinus infections among oyster bars along single tributaries and within single oyster bars. In addition, the study was conducted to compare estimates of prev- alence based on intensive and spatially accurate patent-tong sam- pling with estimates of prevalence based on the traditional dredge sampling of 30 oysters. 381 382 Calvo et al. ^^'^ T^ CHOPTANK RIVER N NANTICOKE RIVER iki/?^ Figure 1. Maryland's portion of the Ciiesapealie Bay showing ORA zones and oyster bar locations In the Chester, Choptank, and Nantlcoke Rivers. A, ORA zone A; B. ORA zone B: C, ORA zone C. Bar code abbreviations: CHOP, Old Field: CHBR, Buoy Rock: CRCA, Cabin Creek; CRMD, Dixon/Mill Dam; CROS, Oyster Shell Point; BCIC, Irish Creek. METHODS A pilot study was conducted during June 1994 to test the use of long range navigation (LORAN) and global positioning system (GPS) navigation for the accurate deployment of patent-tongs for the collection of oysters and to compare methods for P. mariiuis diagnosis by the use of hemolymph and tissue assays. The pilot study was restricted to patent-tong sampling on three oyster bars within the Choptank River — Cabin Creek (CRCA), Dixon/Mill Dam (CRMD), and Irish Creek (BCIC) (Fig. 1; Table 1). In September 1994. the study was expanded to examine the spatial variability of P. marinus infections in other oyster bars within the Chester. Choptank. and Nanticoke Rivers, on the East- em Shore of Maryland (Fig. 1; Table 1). For comparative pur- poses, patent-tong sampUng and dredge sampling were both con- ducted in each of four oyster bars, within 35 d. during September and October 1994 (Table 2). Conflicting schedules of boat time, gear use, and diagnostic services prevented a more simultaneous occurrence of patent-tong and dredge surveys. Dredge sampling was conducted as part of the Modified Fall Survey (MFS), as described below, which is the standard oyster disease-monitoring method in Maryland. Sampling Patent-tong sampling was conducted during June and Septem- ber 1994, in conjunction with site-specific stock assessment sur- veys, in which bars were divided into a series of stations by su- perimposing a grid (0.10 min longitude = 146 m. and 0.05 min latitude = 93 m divisions) to bar delineations. A subset of stations within each bar was selected for sampling on the basis of prior surveys (M. Homer, pers. comm.) indicating the presence of oys- ters within reach of patent-tongs at specific locations. The number of sampling stations per bar ranged from 35 to 155 (Table 1). Oysters were collected with a single patent-tong grab of 1 m" per station. The accurate deployment of patent-tongs at each station was accomplished by LORAN and GPS navigation. The accuracy TABLE L Summarv Field Data. Stations Stations Stations Without Without ToUl Sampled Oysters Oysters Ovsters Bar (N) (N) (%) (N) Spring 1994 CRCA 35 21 60.0 285 CRMD 82 43 52.4 831 BCIC 106 80 75.5 337 Fall 1994 CHBR 38 26 68.4 27 CHOF 155 111 71.6 430 CRCA 35 11 31.4 547 CRMD 49 21 42.8 641 CROS 55 10 18.2 703 NRWS 103 63 61.2 550 Spatial Distribution of P. marinus 383 TABLE 2. Comparison of Prevalence in ORA and MFS Samples. Examined Infected Prevalence Temperature Salinity Bar Dale (N) (N) (%) (°C) (PPt» ORA samples (patent-tong) CHBR 9/08/94 27 4 14.8 23.0 8.0 CHOP 9/08/94 315 57 18.1 23.0 8.0 CROS 9/13/94 528 65 12.3 22.0 6.0 NRWS 9/29/94 487 203 41.7 22.0 13.0 MFS samples (dredge) CHBR 10/12/94 30 3 10.0 16.0 8.5 CHOP 10/12/94 30 6 20.0 16.5 7.5 CROS 10/19/94 30 3 10.0 16.0 8.5 NRWS 10/26/94 30 12 40.0 15.0 8.0 of LORAN and GPS was, respectively. ± 100 and ± 10 m. When fewer than 30 oysters were collected by a grab, all oysters were examined for P. marinus. When grabs contained more than 30 oysters, a random sample of 30 oysters was selected from the pool. The target number of oysters to be collected per bar was 500. The average number of oysters collected per bar, during spring and fall, was, respectively, 484 and 483. Dredge sampling was conducted according to MFS methods established by Smith and Jordan (1993). Following the referred methods, a single sample of 30 oysters was collected per bar with a dredge (91-cm-wide opening). Oysters were randomly selected from an aggregate of five (9-1 substrate) samples collected, each with individual dredge-tows. LORAN navigation was used to help in locating the same approximate site, within each bar, where multiple dredge-tows have been conducted over the years. Tow distance varied depending on the density of the bottom material. Five tows covered a minimum distance of 100 ni. which was necessary to fill the dredge in areas with the highest density of bottom material. All 30 oysters collected in each bar were exam- ined for P. marinus, as described below. Diagnosis In the laboratory, oysters were scrubbed of fouling organisms and shell height was determined for each individual. All sampled oysters were examined for P. marinus by the use of hemolymph Ray fluid thioglycollate medium (RFTM) assays following the technique used at the Cooperative Oxford Laboratory for the rapid diagnosis of oysters (A. Farley, pers. comm.). The technique is a simplification of the hemolymph assay (Gauthicr and Fisher 1990), as described below. To collect hemolymph, a small orifice was drilled into each oyster adjacent to the adductor muscle with a 3.12-mm bit, and a 0.5-ml sample was withdrawn with a 3-ml syringe fitted with a 20-gauge needle. Hemolymph samples were dispensed into 3.4-ml wells of 24-well culture plates (Coming 2582-24) and covered with a 2 ml of RFTM fortified with Chlo- romycetin (0.025 g/ml). A 0.1-ml suspension of mycostatin (500,000 U in 125 ml of distilled water) was then added to each sample. After incubation at room temperature for 7 d, the top layer containing approximately 1 ml of RPTM was carefully aspirated with a transfer pipette and one to two drops of Lugol's iodine solution (6 g of potassium iodide and 4 g of iodine in 100 ml of distilled water) were added to the remaining sample containing oyster hemocytes and parasite cells. Stained hypnospores (pre- zoosporangia) were counted on an inverted microscope. When cell abundance was low (<250 cells), all cells in the well were counted. When cell abundance was higher, cells were counted in five replicate fields of view and the total cell number per well was estimated by multiplying the average cell number per field by the number of fields in the well. If necessary, when cell abundance was very high (>1 million), samples were diluted 10-fold to fa- cilitate counting. The intensity of infection was then ranked, on the basis of cell abundance estimates per I ml of hemolymph, into seven categories with 10-fold increments from stage I (1-10 cells) 1 Percent Prevalence ■>7S% ^51 to 7S X [126 to 60 % D 1to26% X 0% a m n D X n D D CHOP ° n X 0.5 Kilometers Figure 2. Prevalence of P. marinus in Chester River-Old Field (CHOF). Fall 1994. Only sUtions with 10-30 oysters are shown. 384 Calvo et al. CRCA Percent Prevalence B>76% ^61 to 76 % 1126 to SO % D 1to26% X 0% Kilometers To examine geographical distribution of prevalence, data were imported into a geographical information system (Maplnfo 3.0). Because few stations had 30 oysters, stations with 10-30 oysters were used to graphically illustrate the variation of prevalence among stations. Prevalence was divided into categories of 0%, 1-25%, 26-30%, 31-75%, and 76%-100% (Figs. 2-5). To fur- ther examine the variation of prevalence within Nanticoke River- Wilson School (NRWS), the oyster bar was divided into two sub- areas containing contiguous stations. Subarea A contained 13 sta- tions with a total of 145 oysters, and subarea B contained 21 stations and a total of 333 oysters (Fig. 5). Graphical plots of prevalence versus oyster density were used to examine, for stations with 30 oysters, if the two variables were related. To further investigate if oyster density affected P. marinus prevalence, regression analyses were performed (Zar 1984). For regression analysis, prevalence was arcsin transformed to improve normality and homogeneity of variance. A plot of residuals versus predicted values showed no pattern. Nonparametric statistics were used to examine differences in intensity categories for stations with 1-30 oysters. Kruskal-Wallis tests (Zar 1984) were used to examine differences in intensity categories among: (1) tributaries, (2) bars within single tributaries, and (3) stations within single bars. Spearman's rank correlation analysis (Zar 1984) was used to examine the relationship between oyster size and intensity category. A x" analysis was used to Figure 3. Prevalence of P. marinus in the Choptank River-Cabin Creek (CRCA) and Choptank River-Dlxon/Mill Dam (CRMD). Fall 1994. Only stations with 10-30 oysters are shown, to stage VII Ol million cells). The diagnosis of P. marinus was by RFTM assay (Ray 1966, Howard and Smith 1983). Separate RFTM assays were conducted on gill and palp tissue combined and on rectal tissue. Quantitation of low-intensity infections (<250 cells/sample) in combined gill and palp samples and in rectum samples was determined and ranked as in hemolymph sam- ples. Infections of higher intensities (>250 cells/sample) were assigned to categories on the basis of the relative abundance and density of cells, but cell counts were not performed on those samples. Data Analysis Records of individual oysters containing shell height, infection stage, date of sample collection, bar and station codes, and latitude and longitude coordinates were entered into a computer data base. Additional information on oyster density and percent mortality, at the time of sampling, was obtained from stock assessment data bases (M. Homer, pers. comm.). Parasite prevalence was calculated separately for each bar, sta- tion, or size class considered. To avoid unrepresentative data aris- ing from stations with small sample size, only stations with 30 oysters were selected for statistical analysis of prevalence. A x" analysis and a Fisher Exact Test (Zar 1984) were conducted to compare the frequency of infected oysters between bars in the Chester and the Choptank Rivers and between stations within bars. Percent Prevalence B>7S% ^51 to 76 % 026 to 50 % D 1to25% X 0% D D HH mMTl D X XX n CROS 0.5 Kilometers Figure 4, Prevalence of P. marinus in the Choptank River-Oyster Shell Point (CROS). Fall 1994. Only stations with 10-30 oysters are shown. Spatial Distribution of P. marinus 385 NRWS Percent Prevalence ■>76% ■61 to 76 % rnii to 60 % n 1to26% X 0% 0.5 Kilometers Figure 5. Prevalence off. marinus in Nanticolte River-Wilson Shoal (NRWS). Fall 1994. Only stations with 10-30 oysters are shown. A, subarea with ,18.6% prevalence; B, subarea with 72.8% prevalence. compare the frequency of positive parasite detection by rectum, gill, and palp and hemolympli diagnoses in BCIC oysters. RESULTS Spring 1994 The average number of oysters collected from Choptank River bars was 484. Often, no oysters were present in stations, or none were collected by patent-tongs. The frequency of stations where TABLE 3. Prevalence of P. marinus, Oyster Density, Size, and Mortality, Fall 1994. Mean Mean Ovstcr Shell Prevalence Mortality Density Height ± SD Bar (%) (%) (1/m^) (mm) CHBR 14.8 37.2 0.7 96.0(17.2) CHOP 18.2 13.5 2.8 79.0 (26.4) CRCA 2.0 23.5 15.6 57.6(18.2) CRMD 8.2 16.7 13.4 66.5 (14.0) CROS 12,3 18.4 11.3 72.5 (16.6) NRWS 41.7 6.8 5.4 80.4(15.8) oysters were absent, or beyond the reach of patent-tongs, ranged from 52.4'7f in CRMD to 75.5% in BCIC (Table 1 ). At the time of collection, temperature was 26-27°C and salinity ranged from 4 ppt m the upstream bar CRCA to 7 ppt in the downstream bar BCIC. Mean shell height was 59.8 mm in BCIC oysters, 63.0 mm in CRCA oysters, and 70.9 mm in CRMD oysters. On the basis of hemolymph diagnosis, prevalence was 17.2% m CRCA, 3.3% in CRMD, and 89.7% in BCIC. Prevalence among CRCA stations with 30 oysters (N = 3) ranged from 0 to 37%. No prevalence results are presented for stations within CRMD or BCIC, because none of the stations had 30 oysters. Differences in prevalence among Choptank River bars and among stations within CRCA were significant (x" and Fisher Exact sta- tistics, p < 0.05). Most of the infected oysters had light-intensity (stages 1 and II) infections. Pearson's \' statistic indicated that there was no significant (p > 0.05) difference in the prevalence of P . marinus by rectum, gill, and palp and hemolymph diagnoses of the same BCIC individuals (N = 96). Fall 1994 The average number of oysters collected in the Chester, Chop- tank, and Nanticoke Rivers was 483 oysters per bar. A high pro- portion of the stations sampled had no oysters, or none were col- lected by patent-tongs. The percentage of stations with no oysters ranged from 68.4 to 71.6% in Chester River bars; ranged from 18.2 to 42.5% in Choptank River bars; and was 61 .2% in NRWS (Table 1). At the time of sampling, temperature was 22-23°C. Salinity was 8 ppt in Chester River bars, 5-6 ppt in Choptank River bars, and 13 ppt in NRWS (Table 2). Mean shell height ranged from 79.0 to 96.0 mm in Chester River oysters, ranged from 57.7 to 72.5 mm in Choptank River oysters, and was 80.4 mm in NRWS oysters (Table 3). Prevalence among bars, within single tributaries, ranged from 14.8 to 18.1% in the Chester River and from 2.0 to 12.3% in the Choptank River. Prevalence among stations with 30 oysters (within individual bars) ranged from 7.0 to 13.0% in Chester River-Old Field (CHOF), 0 to 10%. m CRCA, 0 to 27% in CRMD, 0 to 30% in CROS, and 50 to 80% in NRWS. Differences in prevalence among Choptank River bars and among stations in CRCA, Choptank River-Oyster Shell Point (CROS), CRMD, and NRWS were significant (x" and Fisher Exact statistics, p < 0.05); differences in prevalence among Chester River bars and among CHOF stations were not significant (x' and Fisher Exact statistics, p > 0.05). Distribution of prevalence among stations, with 10-30 oysters examined, was as follows. Prevalence in most stations located within bars in the Chester and Choptank was 1-25% (Figs. 2-A). Prevalence in most stations within the Nanticoke River bar (NRWS) ranged from 26 to 75% (Fig. 5). However, the difference in prevalence among stations with 30 oysters examined was at most 30%. A pattern of contiguous stations with similar preva- lence ranges was observed in CROS and NRWS (Figs. 4 and 5). Prevalence m NRWS subareas A and B was 38.6 and 72.8%, respectively. Estimates of prevalence derived from intensive patent-tong sampling were remarkably close, within 5%. of estimates of prev- alence based on less intensive dredge sampling (Table 2). On the basis of patent-tong sampling, prevalence for CHBR, CHOF, CROS, and NRWS, was, respectively, 14.8, 18.1, 12.3, and 41 .7%. On the basis of dredge sampling, prevalence for the same bars was, respectively, 10.0, 20.0, 10.0, and 40.0%. 386 Calvo et al. CHBR CHOF HI (14 81%) >IV(1 27%)- m-IV(6 03%)- Ml(10 79%)- NEGAHVE (85 19%) NEGATIVE (81 90%) CRCA |— >IV(0 00%) — IIWV(0 00%) CRMO >IV(0 87%)- III-1V(1 08%)- Hl(6 07%)- NEGATIVE (98 01%) NEGATIVE (91 97%) CROS NRWS >IV(1 89%) lll-IV(4,17%) Ml (6.25%) >IV (3 08%) lll-IV(10 88%) Ml (27 72%) NEGATIVE (58 32%) NEGATIVE (87 69%) Figure 6. Intensity of P. marinus infections in sampled bars in the Cliester, Choptank, and Nanticoke Rivers. Fall 1994. Bar code abbreviations: CHOF, Old Field; CHBR, Buoy Rock; CRCA, Cabin Creek; CRMD. Dixon/Mill Dam; CROS, Oyster Shell Point; BCIC, Irish Creek. Pie charts show percentage of cases in a given category. In general, infection intensities were ligiit. The largest propor- tion of the cases were categorized as 1 (1-100 cells/ml of he- molymph) and II (100-1000 cells/ml of hemolymph) (Fig. 6). Significant differences (Kruskal-Wallis statistic, p < 0.05) in the distribution of intensity categories were found among tributaries; among bars within the Choptank River; and among stations within CHOF, CRMD. CROS, and NRWS bars. Differences among bars in the Chester River and among stations within CHBR and CRCA were not significant (Kruskal-Wallis statistic, p > 0.05) (Table 4). Overall, most of the infected oysters collected in the fall were large (50-200 mm). Prevalence in small (20-50 mm) oysters was 9.9% (38 infected in 383 examined), and prevalence in large oys- ters was 20.8% (456 infected in 2,187 examined). Thus, the ratio of the percentage of small infected oysters to large infected oysters was roughly 1;2. Analysis of data for individual bars indicated that the referred 1:2 ratio held for bars with low salinity (e.g., CHOF) and high salinity (e.g., NRWS). The relationship between oyster size and infection intensity was also investigated with the whole fall data set (N = 2,624 observations) and subsets of data (N = 27-528 observations) for ndividual bars. The following results apply to both the whole data set and individual subsets of data. A plot of size versus infection stage revealed that stages 4 and above were less frequent in small (20- to 50-mm) oysters and very large (100- to 200-mm) oysters than in medium (50- to 100-mm) oysters. The correlation between TABLE 4. Comparison of P. marinus Infections Among ORA Tributaries, Bars, and Stations. K-S Comparison Statistic DF p Value Significance Among tributaries 287.05 2 0.0001 S Among bars within Chester River 0.27 1 0.5999 NS Among bars within Choptank River 34.80 ") 0.0001 S Among stations within CHBR 15.78 13 0.2613 NS Among stations within CHOF 75.45 42 0.0012 S Among stations within CRCA 22.71 25 0.5946 NS Among stations within CRMD 52.36 27 0.0024 S Among stations within CROS 92.17 42 0.0001 s Among stations within NRWS 77.51 36 0.0001 s Fall 1994. Results of Kruskal-Wallis (K-S) tests on oyster hemolymph infection stages. At a = 0.05: S. significant; NS, not significant. Spatial Distribution of P. marinus 387 oyster size and infection stage, however, was not significant (Spearman's rank correlation statistic, p > 0.05). Percent oyster mortality ranged from 13.5 to 37.2% in Chester River bars; ranged from 16.6 to 23.5% in Choptank River bars; and was 6.7% in NRWS. Mean oyster density ranged from 0.7 to 2.8 oysters/m" in Chester River bars; ranged from 11.3 to 15.6 oysters/m" in Choptank River bars; and was 5.4 oysters/m* in NRWS. Mean oyster density in NRWS subareas A and B was 1 1.5 and 18.1 oysters/m". respectively. Prevalence, percent mortality, and mean oyster density did not show corresponding patterns among oyster bars (Table 3). Prevalence was not affected by den- sity (R- 0.011. p < 0.05). DISCUSSION In agreement with MFS results, this study showed that P. mari- nus infections in Maryland ORAs were of low prevalence and intensity during 1994. MFS results indicated that the prevalence of P. marinus. for most bars sampled in fall 1994, was 30% below corresponding values for 1993 and 1992. Haplosporidium netsoni was absent from most samples examined during the MFS con- ducted in 1994 (Krantz 1995). The low prevalence and intensity of P. marinus infections, and the near absence of H. nelsoni infec- tions, during 1994 can be attributed to extreme low-salinity con- ditions prevailing in the Chesapeake Bay during June to September 1994. when mean streamflow into the bay was above the long- term (>60 y) average (U.S. Geological Survey 1994, Krantz 1995). This study showed that there was no significant difference (Pearson's x". P > 0.05) in the frequency of P. marinus detection (prevalence) by rectum, gill, and palp and hemolymph diagnoses of the same individual oysters. Similarly, Bushek et al. (1994) reported no significant differences in P . marinus prevalence as determined by hemolymph and rectum assays. Seasonal, spring to fall, comparison of P. marinus prevalence in Choptank River bars showed an increase from 3.3 to 8.2% in CRMD but a decrease from 17.7 to 2.0% in CRCA. The seasonal increase in prevalence, in CRMD, was expected on the basis of the characteristic seasonal cycle of disease progression with increasing temperatures from spring to fall (Andrews 1988). The unexpected seasonal decrease in prevalence in CRCA may have been related to a long duration (3 mo) of low salinity (4-5 ppt, at the time of spring-fall sampling at that location). Reduction of infection prev- alence during a period of low salinity and warm temperature in CRCA may be attributed to a disproportionally high mortality of infected oysters stressed by extreme low-salinity conditions, as described above. Extreme low-salinity conditions during 1994 may have been conducive to relatively high oyster mortality in CRCA (23.5%) and CHBR (37.2%). On the basis of very limited mortality data, restricted to once-yearly box counts conducted in the fall, we speculate that the benefit afforded to oysters by low- salinity environments as a refuge from the presence and virulence of parasites may have been offset in some ORAs during 1994 by the adverse effects of extreme low salinity on oyster survival. As expected, the geographic distribution of prevalence in sam- pled ORAs corresponded with the salinity regime of the particular area. Oyster bars located in areas with low salinity had low par- asite prevalence, and oyster bars located in areas with high salinity had high parasite prevalence, in most cases, variation in preva- lence among oyster bars within sampled ORA tributaries and among stations within oyster bars was statistically significant but small (<30%'). In the spring, however, prevalence between the upstream and the downstream Choptank River bars (CRCA and BCIC) varied by as much as 87% in an area separated by 40 km of distance and with salinity within 3 ppt. Unfortunately, it was not possible to determine if spring prevalence in BCIC (89.7%) was maintained in the fall, because BCIC was substituted with CROS during fall sampling. The difference in fall prevalence between CRCA and CROS was only 10%, but CRCA and CROS had salinity within 1 ppt and were only 7 km apart. Ray (1987) re- ported that very large differences in prevalence (96%) were main- tained for 1 y between monthly samples collected from two oyster reefs within 10 km and subjected to similar environmental condi- tions. Ray's study continued for an additional year, when infec- tions intensified and the referred difference was eventually re- duced to a minimum of 4%. The existence of relatively large differences in P . marinus infections among oyster bars located in nearby areas and experiencing similar environmental conditions may be related to the close proximity of oysters required for the transmission of P. marinus (Andrews 1965, Andrews 1988, Ray 1987). Mackin (1962) suggested that large doses of P. marinus cells, mostly originating from decomposing tissues in dead oys- ters, are the main source of new infections for surrounding oysters. The role of water flow in transmission dynamics (Mackin 1962, Ray 1987) remains unclear. However, water currents may con- centrate or disperse parasites and favor or discourage the estab- lishment of foci of infections, depending on the pattern of the How. Investigations on water flow patterns around oyster bars may be necessary to better explain parasite dispersal and the spatial distribution of infections. Prevalence variability among stations with 30 oysters, within a bar, was low (<30%). However, this variability increased when sample size was reduced to include 10-30 oysters per station, even though the total number of oysters and stations available for com- parison increased. For instance, prevalence in NRWS stations (N = 22) with 10-30 oysters (N = 404 oysters) varied by as much as 72%. and prevalence in NRWS stations (N = 3) with 30 oysters (N = 90 oysters) varied by 30%. Because there were few stations with 30 oysters, it was necessary to include stations with 10-30 oysters to examine the spatial variation distribution of prevalence. However, a sample of 30 or more oysters is required to statistically detect a minimum of 10% prevalence in a population of 100.000 oysters (Amos 1985). Stations with 30 oysters in NRWS. how- ever, may have been located in areas having higher prevalence than stations with 10 oysters because the overall prevalence within NRWS was 41.7% (N = 487 oysters) and the prevalence for stations with 30 oysters (N = 90 oysters) was 63.3%. By parti- tioning NRWS into subareas containing a series of contiguous stations, it was possible to determine that prevalence in the subarea containing stations with 30 oysters (N = 333 oysters) was 72.8% and prevalence in the other subarea (N = 145 oysters) was 38.6% (Fig. 5). Differences in prevalence among stations, within most bars, were significant. This result appears to contradict Craig et al. (1989). who did not find significant differences (binomial test, p > 0.05) in prevalence within bars, after comparing 49 bars with three stations in each bar. In this study, there were significant differences (x" and Fisher Exact Tests, p < 0.05) in prevalence among stations (N = 35-155) in four of five of the oyster bars examined. Comparison of our results with those of Craig et al. is complicated, however, because the referred authors did not spec- ify the type of gear used in a particular location other than indi- cating that hand, tong, or dredge was used, depending on water 388 Calvo et al. depth. White et al. ( 1989) reported that, in general, infected oys- ters were negatively autocorrelated (evenly) distributed in two reefs in Texas, even when oysters were positively correlated (dis- tributed in patches) in one of the reefs. However, White et al. also indicated that infected oysters were distributed in patches at spatial scales >60 cm at both reefs sampled. White et al. used a 50-cm- wide by &50-cni-long rectangle (with its length adjusted accord- ing to oyster density) to sample 60 oysters per reef. By examina- tion of correlograms with test statistic plotted versus distances between one to seven clumps of oysters (12-84 cm). White et al. were able to detect patches of infected oysters at the scale of 60-84 cm. but not at <60 cm. By the use of different methods, this study showed that groups of stations with similar parasite prevalence occurred (in patches) over areas at scales of 100-500 m (Fig. 5). It appears that the distribution of P. mariims within bars was more sporadic in oyster bars with low prevalence (e.g.. CRCA and CRMD) than in oyster bars with higher prevalence (NRWS). Craig et al. (1989) examined the regional distribution of P. marinus in Gulf Coast oyster beds and found that prevalence was positively autocorrelated (an indication of patchiness) at scales of =s300 and 5^1.500 km. Comparing the distribution of infected oysters at different scales, however, may not be valid. Often in aquatic eco- systems, the scale at which heterogeneity manifests itself varies widely and changes of scale can transform a heterogeneous pattern into a homogeneous one and vice versa (Dutilleul 1993). Prevalence estimates based on the patent-tong collection of an average of 340 oysters per bar were in agreement with prevalence estimates based on dredge samples of only 30 oysters per bar (MFS). In most oyster bars examined, estimates of prevalence based on dredge samples taken in September corresponded with slightly (<5%) lower estimates based on patent-tong samples taken in October. Decreasing water temperatures in October. 6°C lower than in September, may have contributed to the small de- crease in prevalence. We speculate that estimates of prevalence based on dredge sampling reflect estimates of prevalence based on patent-tong sampling because dredge samples were composed of oysters collected over a relatively large area. Thus, spatial vari- ability in prevalence was masked by dredge sampling, and intense patent-tong sampling was necessary to show significant, albeit small (<30%), differences in prevalence among stations. As expected. P. marinus infections were more prevalent in large (50- to 200-mm) oysters than in small (20- to 50-mm) oys- ters. It has generally been accepted that P. marinus infections in small oysters are less prevalent and intense than those in larger oysters. White et al. (1989) found infections in large (>50-mm) oysters to be three to four times as prevalent as infections in small (<50-mm) oysters. We found infections in large (50- to 200-mm) oysters to be twice as prevalent as infections in small (20- to 50-mm) oysters. Ray ( 1953) found oysters of increasing age to be increasingly susceptible to infections, and Paynter and Burreson (1991) noted that larger animals tended to be infected sooner and more intensely than smaller ones. Small oysters or spat may be less susceptible to infections, given the limited volume of water they filtered compared with large animals (Paynter and Burreson 1991). Surprisingly, however, oyster size and infection intensity were not correlated in this study. The lack of correlation between oyster size and infection intensity may be attributed to the overall scarcity of higher than light intensity infections as compared with abundant light infections. Average oyster density and prevalence revealed no correspond- ing patterns. However, there were only minor differences in mean oyster density among bars. Perhaps, densities within the range examined (0.7-15.6 oysters/m") did not affect disease acquisition and/or transmission dynamics. Density varies widely within bars, however, and the difference in prevalence between subareas in NRWS corresponded with a difference in oyster density. A rela- tively low prevalence (38.6%) corresponded with a relatively low oyster density (11.5 oysters/m~) in NRWS subarea A. and a rel- atively high prevalence l72.8'7c) corresponded with a relatively high oyster density (18.1 oysters/m") in NRWS subarea B. It is also interesting to note that the parasite was present even at host densities less than I oyster/m". Therefore, a very low oyster den- sity may be necessary to avoid disease transmission. The relationship between prevalence and mortality was not clearly defined. On the basis of very limited mortality data, as previously described, we make the following observations. The pattern of prevalence did not correspond with the pattern of mor- tality (Table 3). In fact, the bar with the highest prevalence (NRWS with 41 .7%) had the lowest mortality (6.8%). the bar with the lowest prevalence (CRCA with 2.0%) had 23.5% mortality, and CHBR. with a prevalence of only 14.8%. had the highest mortality (37.2%). Mortality estimates, however, are based on annual counts of live and dead oysters, which may correspond to lethal infections present during the year before sampling. Thus, high prevalence levels during 1993 (roughly two to four times higher than in 1994) may have resulted in disease-induced mor- tality during 1994. In summary, results from this investigation support three con- clusions. (1) During fall 1994, variation of P. moW/iMi prevalence, among and within oyster bars in sampled ORA tributaries, was small (<30%) but statistically significant. (2) Traditional dredge samples of 30 oysters per bar provided estimates of prevalence remarkably similar to the ones obtained from patent-tong samples of an average of 340 oysters per bar. (3) RFTM assays conducted in the spring showed that hemolymph. rectum, and combined gill and palp samples gave equivalent determinations of prevalence. ACKNOWLEDGMENTS This work was supported by federal grant #NA47FL0I55 of the National Oceanic and Atmospheric Administration of the De- partment of Commerce. We thank Dr. Marc Homer and his staff for assistance with patent-tong sampling, and the staff at the Co- operative Oxford Laboratory for assistance with oyster disease diagnosis. We appreciate editorial comments and suggestions by an annonymous reviewer and by Dr. George Krantz. LITERATURE CITED Amos, K. H. 1985. Procedures for the detection and identification of certain fish pathogens. 3rd ed. Fish Health Section. Amencan Fishenes Society. Corvallis, Oregon. Andrews. J. D. 1965. Infection experiments in nature with Dermo- cystidium marinum in Chesapeake Bay. Chesapeake Sci. 6(1):60- 67. Spatial Distribution of P. marinvs 389 Andrews. J. D. 1988. Epizootiology of the disease caused by the oyster pathogen Perkinsus marinus and its effects on the oyster industry. Am. Soc. Spec. Publ. 18:47-63. Andrews, J. D. & S. M. Ray. 1988. Management strategies to control the disease caused by Perkinsus marimis. Am. Soc. Spec. Publ. 18:257- 264. Bushek. D., S. E. Ford & S. K. Alien. Jr. 1994. Evaluation of methods using Ray's fluid thioglycollate medium for diagnosis of Perkinsus marinus infection in the eastern oyster. Crassoslrea virginica. Annu. Rev. Fish Dis. 4:201-207. Craig, A., E. N. Pail, R. R. Fay & J. M. Brooks. 1989. Distribution of Perkinsus marinus in Gulf Coast Oyster populations. Estuaries 10(2): 82-91. Dutilleul. P. 1993. Spatial heterogeneity and the design of ecological field experiments. Ecology 74(6): 1646-1658. Ford. S. E. 1992. Avoiding the transmission of disease in commercial culture of molluscs, with special reference to Perkinsus marinus (Dermo) and Haplosporidium nelsoni (MSX). J. Shellfish Res. 1 1(2): 539-546. Gauthier. J. D. & W. S. Fisher. 1990. Hemolymph assay for diagnosis of Perkinsus marinus in oysters Crassoslrea virgiiuca (Gmelin. 1791 1. J Shellfish Res. 9(2):-^67-371 . Howard. D. W. & C. S. Smith. 1983. Histological Technique for Marine Bivalve Mollusks. NOAA Technical Memorandum NMFS-F/NEC-25. U.S. Department of Commerce, Washington. D.C. Krantz. G. 1995. Maryland Oyster Population Status. 1993 and 1994 Biological Seasons. Fisheries Division. Maryland Department of Nat- ural Resources. Annapolis. Maryland. Mackin. J. G. 1962. Oyster disease caused by Dermocysliduim marinum and other microorganisms in Louisiana. Te.xas Institute of Marine Sci- ence Publication 7:132-229. Maryland Oyster Roundlable Action Plan. 1993. Annapolis. Maryland. 26 PP- Paynter. K. T. & E. M. Burreson. 1991. Effects of Perkinsus marinus infection in the eastern oyster, Crassoslrea virginica: 11. Disease de- velopment and impact on growth rale at different salinities. J. Shellfish Res. I0(2):425-431. Ray. S. M. 1953. Studies on the occurrence oi Dermocysiidium marinum in young oysters. Proc. Nail. Shellfish. Assoc. 44:80-92. Ray, S. M. 1966. A review of the culture method for detecting Dermo- cysiidium marinum. with suggested modifications and precautions. Proc. Nail. Shellfish. Assoc. 54:55-69. Ray. S. M. 1987. Salinity requirements of the American oyster, Crassos- lrea virginica. pp. E1-E28. In: A. J. Mueller and G. A. Matthews (eds.). Freshwater Inflow Needs of the Matagorda Bay System With Focus on Penaeid Shnmp. NOAA Tech. Mem. NMFS-SEFC-189. U.S. Department of Commerce. Washington. DC. Smith. G. & J. Jordan. 1993. Monitoring Maryland's Chesapeake Bay Oysters. Maryland Department of Natural Resources. CBRM-OX-93- 3. Chesapeake Bay Research and Monitoring Division. Annapolis, Maryland. 93 pp. U.S. Geological Survey. 1994. Estimated Streamflow Entering Chesa- peake Bay. Towson, Maryland. White. W. E.. E. N. Pail. E. A. Wilson & S. M. Ray. 1989. The spatial distribution oi Perkinsus marinus. a protozoan parasite, in relation to its oyster host {Crassoslrea virginica) and an ectoparasitic gastropod [Boonea impressa). J. Mar. Biol. Assoc. U.K. 69:70,^-717. Zar, J. H. 1984. Biostatlstical analysis. 2nd cd, Prenticc-Hall. Inc.. En- glewood Cliffs, New Jersey. Joiirmil ot Shellfish Research. Vol. 1?, No. 2. 341-344. 19%. CORRELATION BETWEEN THE PRESENCE OF LATHYROSE WITH THE ABSENCE OF HAPLOSPORIDWM NELSONI IN CRASSOSTREA VIRGINICA FROM TWO SOUTH CAROLINA TRIBUTARIES WHERE PERKINSUS MARINVS ALSO INHIBITS HEMOCYTE AGGLUTINATION BY THE LATHYRUS ODORATUS LECTIN THOMAS C. CHENG'* AND JOHN J. MANZl'^ ^Shellfish Rt'Si'circh Instiluie and ^SeaPerfect Atlantic LittleNeck ClaniFanns P.O. Bo.x 12139 Charleston, South Carolina 29422 ABSTRACT Hemocytes from Crassoslrea virginica from two adjacent tributaries in South Carolina were exposed to six dilutions of the Lalhxnis odoiarus lectin. As a result, it has been further confirmed that there is a correlation between the presence of lathyrose, the yet-uncharacterized saccharide that binds to this lectin, and the essential absence of the pathogen Haplosporidnim nelsoni. Furthermore, it has been demonstrated that there was Inhibition of the agglutination of lathyrose-positive hemocytes by the L. odoratus lectin when a second protislan pathogen, Perkinsiis marimu. was present. This confirms that lathyrose (or a functionally very similar molecule) occurs on the surface of P. mariniLs. Comparisons of the H nelsoni infection frequencies in oysters from the two sites suggest that the intertidal marsh separating the two sites was a sufficient barrier to result in shifts in infection frequencies within 12 mo. KEY WORDS: Crasso.slrea viri>inica. Haplosporidiiim nelsoni. Perkinsiis marimis. lathyrose. L. odoratus lectin INTRODUCTION It is known that the oyster pathogen Haplospondium nelsoni occurs in coastal South Carolina (Dougherty et al. 1993). Subse- quent surveys by personnel of the South Carolina Marine Re- sources Research Institute (SCMRRI) (unpubl.) have revealed that — 25% of the oysters. Crassoslrea virginica. (Gmelin), collected during May through September 1994. from Inlet Creek and Toler's Cove, tributaries of Charleston Harbor and the Atlantic Ocean, respectively, on Sullivan's Island, Charleston County. South Carolina, harbored this parasite. A second protistan parasite. Per- kinsiis marinus. also occurs in oysters at these sites. Because coastal South Carolina is punctuated by numerous tributaries and minor inlets commonly separated by intertidal marshes, the principal question being posed was whether there is transmission of infections by such protistan parasites as H. nelsoni between oysters native to two adjacent tributaries during a season when the separating marsh was essentially dry. An answer to this question could shed some light on the transmission mechanism of H. nelsoni. i.e.. whether a continuous aquatic milieu is essential for transmission. Becauseearlier studies (Cheng et al. 1994a. Cheng et al. 1995) have revealed that there is a correlation between the occurrence of a yet-uncharacterized saccharide designated as lathyrose (Cheng and Dougherty 1994a) on the surface of hemocytes from oysters free of//, nelsoni, a second objective of the study being reported here in was to ascertain whether there is a similar correlation between the presence of lathyrose and the absence of//, nelsoni in oysters from adjacent tributaries that had revealed a similar inci- dence of infection with //. nelsoni. Third, because earlier studies (Cheng and Dougherty 1994b. Cheng and Dougherty 1995) had suggested that lathyrose occurs on the surface of intramolluscan stages of P. marinus in Chesa- peake Bay and Apalachicola Bay oysters, we were interested in obtaining additional evidence to support or reject the commonality of this phenomenon by testing P. marimis from oysters from es- sentially the same location. MATERIALS AND METHODS Oysters *Corresponding author. A total of 65 oysters were collected from Inlet Creek during March through September 1995. These specimens measured from 69.1 to 111.3 mm in length. Sixty-eight oysters were collected from Toler's Cove during the same period. These measured from 59.9 to 140 mm in length. Inlet Creek and Toler's Cove, separated by approximately 1.5 miles, are connected by an intertidal marsh that is commonly completely devoid of water during a dry season, as during the summer of 1995. The salinity at both collection sites during the collection period was 22-28%c. Hemolymph Collection After the external surfaces of the oysters were cleaned, approx- imately 2 ml of whole hemolymph were collected from the adduc- tor muscle sinus of each oyster by use of a sterile 21 -gauge hy- podermic needle and a 1-ml tuberculin syringe. The samples were washed three times in isotonic (540 mOsm) saline involving cen- trifugation at 300g. After the third wash, the cell pellets were gently resuspended in 2 ml of isotonic saline. The final cell counts averaged 1-2 x 10-^/ml. Lectins The Lallnrtis odoratus (sweet pea) lectin was tested against washed oyster hemocytes. The initial lectin solution used was at a concentration of 0. 1 mg/ml. It was prepared in phosphate-buffered saline, pH 7.4, and was serially diluted twofold with isotonic saline in microtiter plates to give final dilutions of 1:1 to 1:2,048. It is known that the agglutination of oyster hemocytes by the L. odoratus lectin is not inhibited by A'-acetyl-D-glucosamine. D( + )- glucose, or d( + )-mannose. the known inhibitors of cell aggluti- 391 392 Cheng and Manzi nation mediated by this lectin ( Ticha et al . 1 980 ) ; nevertheless . the possible inhibitory effects of two of these saccharides. N-acetyl- D-glucosamine and d( + )-glucose, in the hemocyte-lectin combi- nations comprising this study were tested. In addition to the L. odoratus lectin, the Conavalia ensiformis lectin (Con A. type III. jackbean) was also tested against washed oyster hemocytes. These tests served as positive controls because it is known that Con A will agglutinate hemocytes of C. virginica (Yoshino et al. 1979. Cheng et al. 1980. Cheng et al. 1993, Kanaley and Ford 1990). In the tests involving Con A. the con- centration of the initial lectin solution was 1.0 mg/ml and W-acetyl-D-glucosamine and d( -I- )-glucose were used as the inhi- bition saccharides. The Con A solution was prepared in phosphate- buffered saline and was serially diluted twofold with isotonic sa- line in microtiter plates to give final dilutions of I ;1 to 1:2.042. In negative control tests, isotonic saUne instead of lectin was used. None of these resulted in the agglutination of hemocytes. In inhi- bition tests involving both the L. odoratus lectin and Con A. each lectin was diluted serially in 0.2 M solutions of the appropriate inhibition sugar. Agglutination tests were carried out in 96-well U-bottom plates (Cell Wells. Coming. NY). Fifty microliters of hemocyte suspen- sion was added to each experimental and control well, and the plates were incubated for 24 h at room temperature (25°C). Earlier smdies (Cheng and Dougherty 1994a. Cheng et al. 1993, Cheng et al. 1994. Cheng et al. 1995) had revealed that not all of the hemocytes exposed to the L. odoratus lectin or Con A agglutinated. Therefore, the percentages of clumped cells (i.e.. number of agglutinated cells; 100 cells) were ascertained at the highest concentration of the two lectins tested as well as at dilu- tions of I;2. 1;64. 1:512. 1:1.024. and 1:2.048. The counting of agglutinated and single cells was achieved microscopically. When three or more cells were clumped, these were considered to be agglutinated. Pairs of cells were seldom observed. Detection of Infections Determinations of the presence or absence of H. nelsoni and/or P. marinus were carried out in two ways: ( 1 ) by histological ex- amination of representative hematoxylin & eosin-staincd sections. and (2) by using a modification of the panning technique of Ford et al. (1990). which Cheng and Dougherty (1995) have reported to be as effective qualitatively for the detection of protistan parasites in oysters as the hemolymph assay method of Gauthier and Fisher (1990). Also, as Bushek et al. (1994) have pointed out. the com- monly used fluid thioglycollate medium used for the qualitative diagnosis of P. marinus in C. virginica has major drawbacks. Consequently, we elected to use a combination of the histological and panning techniques in our determination of the possible pres- ence of P. marinus and also of H. nelsoni. It is possible that by using these methods, we may have overlooked some very light infections with P. marinus. In brief, the panning technique involves placing ~ I ml of whole hemolymph on the bottom of a Petri dish and permitting it to stand for 30 min at 25°C. Because oyster hemocytes portray greater adherence to the substrate, microscopical examination of nonadhering cells permits the identification of H. nelsoni and/or P. marinus. The tissues prepared for histological examination and the hemolymph samples subjected to panning were from the same oysters from which the hemocytes used for the lectin studies were obtained. Statistical .Analyses To test for significance between the difference in percentages of agglutinated hemocytes from oysters infected with H. nelsoni or P. marinus and between uninfected and infected oysters from both collection sites that had been exposed to the dilutions of the L. odoratus lectin tested, the two-sample t test (Neter et al. 1983) was used. Furthermore, the results were verified by the use of Bonfer- roni's correction for multiple t tests (Neter et al. 1983). RESULTS Inlet Creek Oysters Presented in Table I are the percentages of agglutinated hemo- cytes ± standard deviations (SD) from Inlet Creek oysters that were infected with either H. nelsoni or P. marinus. as well as those of hemocytes from uninfected oysters. All three categories of hemocytes had been exposed to six dilutions of the L. odoratus lectin. The results of inhibition tests involving N-acetyl-D- glucosamine and D{ + )-glucose are also presented. Toler's Cove Oysters Presented in Table 2 are the percentages of agglutinated hemo- cytes ± SD from Toler's Cove oysters that were infected with Mean Percentages of Clumped Oyster Hemocytes TABLE \. SD From Inlet Creek That Had Been Exposed to Six Dilutions of the L. odoratus Lectin. Lectin Dilul ions Group 1:1 1:2 1:64 1:512 1:1,024 1:2,048 Infected with H. nelsoni (n = 12) 4.4 ± 0.2 0 0 0 0 0 Infected with P irwrinii.'i (n = 19) 59.5 ± 5.8 45.0 ± 4.2 43.5 ± 4.4 9.2 ± 2.3 2.3 ± 1.5 1.2 ± 1.1 Uninfected (n = 34) 79.4 ± 8.3 52.3 ± 6.4 48.5 ± 7.1 38.4 ± 6.6 24.3 ± 4.9 13.3 ± 1.7 Inhibition saccharides jV-Acetyl-D-glucosamine ninh ninh ninh ninh ninh ninh D( 4 )-Glucose ninh ninh ninh ninh ninh ninh The oysters from which the hemocyte samples were collected were infected with either H. nelsoni or P. marinus or were not infected. The results of inhibition tests involving two saccharides are also presented, ninh. no inhibition. LaTHYROSE and H. NELSONl AND P. MARINUS IN SOUTH CAROLINA OySTERS 393 Mean Percentages of Clumped Oyster Hemocytes TABLE 2. SD From Toler's Cove That Had Been Exposed to Six Dilutions of the L. odoratus Lectin. Origin of Oysters Lectin Dilutions Toler's Cove 1:1 1:2 1:64 1:512 1:1,024 1:2.048 Infected with H nelsoni (n = 19) 6 5 ± 1.2 5.1 ± 0.4 0 0 0 0 Infected with P. marinus (n = 25) 78.0 ± 7.2 63.8 ± 5.3 35.1 ± 8.4 27.5 ± 2.4 18.3 ± 5.2 4.1 ± 2.1 Uninfected (n = 24) 84.7 ± 7.3 73.7 ± 6.9 52.2 ± 9.4 42.0 ± 8.3 29.8 ± 9.4 17.0 ± 3.3 Inhibition saccharides N-Acetyl-D-glucosamine ninh ninh ninh ninh ninh ninh D( + )-Glucose ninh ninh ninh ninh ninh ninh The oysters from which the hemocyte samples were collected were infected with either H. nclsoiii or P. marinus or were not infected. The results of inhibition tests involving two saccharides are also presented ninh. no inhibition. either H. nelsoni or P. marinus and those of uninfected oysters from the same location. All of the hemocytes had been exposed to six dilutions of the L. odoralus lectin. The results of inhibition tests involving A'-acetyl-D-glucosamine and d( + )-glucose are also presented. Histological Sections Vs. Panning As stated, the detection of H . nelsoni and/or P . nianinis was carried out in two ways: ( 1 ) by histological examination of repre- sentative hematoxylin & eosin-stained sections, and (2) by using a modification of the panning technique of Ford et al. (1990). By comparing qualitative results, i.e.. the presence or absence of either one or both of the pathogens, exact correspondence was ascertained. The numbers of oysters infected with H . nelsoni or P . marinus from both collecting sites are presented in Tables 1 and 2. Doubly infected oysters were not found. DISCUSSION H. nelsoni Infections As presented in Table 1, hemocytes from Inlet Creek oysters that harbored H. nelsoni were not agglutinated at most of the concentrations of the L. odoratus lectin tested. The only lectin concentration at which hemocyte agglutination occurred was at the 1:1 dilution. Even then, of the 12 hemocyte samples tested, only 2 revealed very few agglutinated cells, hence, the very low mean percentage (4.4 ± 0.2%) of clumped cells. The difference be- tween 4.4 ± 0.2 and 79.4 ± 8.3% (the mean percentages of agglutinated cells from H . nelsoni-miscied and uninfected oysters at the 1:1 dilution of the L. odoratus lectin) is significant (P < 0.001). As also presented in Table 1 . some hemocytes from Inlet Creek oysters infected with P . marinus agglutinated when exposed to all six concentrations of the L. odoratus lectin. Furthermore, as ex- pected, the mean percentages of agglutinated cells diminished as the dilution of the lectin was increased. Statistical comparisons of the percentages of clumped cells from uninfected oysters with those from P . m(7n/iw.9-infected oysters revealed that the lower mean percentages associated with infected oysters was significant (P< 0.001) when exposed to 1:1, 1:512, and 1 : 1 ,024 dilutions of the L. odoratus lectin. Neither N-acetyl-D-glucosamine nor D( -I- )- glucose inhibited the agglutination of hemocytes by the L. odora- tus lectin (Table I ). In the case of H. nelsoni-miccieA oysters from Toler's Cove, a small percentage (6.5 ± 1 .2 and 5. 1 ± 0.4% ) of their hemocytes agglutinated when exposed to the 1:1 and 1:2 dilutions of the L. odoratus lectin (Table 2). Moreover, among the 19 H. nelsoni- infected oysters, only a small number of hemocytes from 3 agglu- tinated when exposed to the two lowest lectin dilutions tested. The differences between the percentages of clumped cells from unin- fected oysters and those from oysters infected with H. nelsoni are significant (P < 0.001). Also presented in Table 2 are the percentages of agglutinated hemocytes from oysters with P. marinus when exposed to six concentrations of the L. odoralus lectin. Statistical comparisons between the mean percentages of clumped cells from P. marinus- infected and uninfected oysters revealed that the percentages of clumped cells are significantly lower (P < 0.001 ) when exposed to the 1:2. 1:64. 1:512. and 1:2.048 dilutions of the L. odoratus lectin. Neither A'-acetyl-D-glucosamine nor D( + )-glucose inhib- ited the agglutination of hemocytes exposed to the lectin (Table 2). Results Kith Con A As expected, in our positive controls, hemocytes from the three categories of oysters (uninfected, infected with H. nelsoni. in- fected with P. marinus) that had been exposed to the six dilutions of Con A agglutinated. Moreover, the percentages of clumped cells decreased as the lectin dilution increased (data not shown but essentially identical to those presented by Cheng et al. 1994). The agglutination of hemocytes exposed to Con A at all six dilutions was inhibited by A'-acetyl-D-glucosamine and D( + )-glucose. In view of the above, it may be concluded that the earlier observations that there is a correlation between the occurrence of lathyrose on the hemocyte surface of C . vir^inica and the essential absence of H . nelsoni appear to be supported, in this case, in oysters from the same region. Furthermore, the statistically insig- nificant difference between the mean percentages of agglutinated hemocytes in uninfected oysters from Inlet Creek and Toler's Cove (79.4 ± 8.3 and 84.7 ± 7.3%) (Tables 1 and 2) suggests that oysters from these two adjacent tributaries probably belong to the same subpopulation. However, among the oyster samples exam- ined (65 Inlet Creek and 68 Toler's Cove oysters), 12 (18.5%) were infected with H . nelsoni and 19 (29.2%) were infected with 394 Cheng and Manzi P. marinus at Inlet Creek and 19 (27.9%) were infected with H. netsoni and 25 (36.8%) were infected witbi P. marinus at Toler's Cove. These differences in infection frequencies at the two sites suggest that the 1 .5 miles of intertidal marsh (which is periodically dry) was sufficient as a barrier to result in the stated differences, i.e., the pathogens are totally or partially invagile. As stated earlier, personnel of the SCMRRI, as a result of a survey conducted during May through September 1994, reported that -25% of the oysters from Inlet Creek and Toler's Cove har- bored H. nelsoni. Thus, it would appear that during a 12-mo period, H. nelsoni infection frequencies can be altered as a result of invagility. Relative to the third question posed, i.e., whether there is inhibition of hemocyte agglutination by the L. odoratus lectin by intramolluscan stages of P. marinus. the significantly lower per- centages of clumped cells from P . nmrinus-mtecXtA oysters from both Inlet Creek and Toler's Cove when compared with the per- centages of clumped cells in uninfected oysters (Tables 1 and 2) indicate that inhibition did occur. This supports the earlier reports (Cheng and Dougherty 1994b, Cheng and Dougherty 1995) that such occurs and that the basis is possibly the result of the presence of lathyrose on the surface of P. marinus. Thus, it is predicted that in areas where H . netsoni and P . marinus coexist, one would not expect to find the high percentages of agglutinated hemocytes when exposed to the L. odoratus lectin that exist in areas where P. marinus does not occur. Furthermore, if lathyrose is indeed in some yet undetermined way associated with the innate resistance of C. virginica to H nelsoni. then the presence of P. marinus may influence the susceptibility of C. virginica to H. nelsoni. Finally, our finding of the exact correspondence between the prevalences of W. nelsoni and P. marinus in oysters from both Inlet Creek and Toler's Cove examined by histological study and panning suggests that panning is a reliable qualitative diagnostic technique. .ACKNOWLEDGMENTS The technical assistance of Nancy Pollard and Jill Johnson of the Shellfish Research Institute, Donnia Richardson of the South Carolina Marine Resources Research Institute, and Dr. Hurshell H. Hunt, Department of Biostatistics, Epidemiology and Systems Science of the Medical University of South Carolina, in statistical analysis, is gratefully acknowledged. This research was supported by a grant (No. 94-60038) from the National Science Foundation. LITERATURE CITED Bushek, D., S. E. Ford & S. K. Allen, Jr. IW4. Evaluation of methods using Ray's fluid thioglycollate medium for diagnosis of Perkmsus marinus infection in the eastern oyster. Crassostrea virginica. Aiiiiii Rev. FishDis. 4:201-217. Cheng, T. C. & W. J. Dougherty. I9y4a, Occurrence of "lathyrose" on hemocytes of Haplosporidium nelsoni-free oysters from Chesapeake Bay, USA. Res. Rev. Parasit. 54:117-120. Cheng. T. C. & W. J. Dougherty. I')94h. Inhibition of hemocyte agglu- tination by Lallnrus odoralus lectin in Chesapeake Bay. USA. oysters (Crassostrea virginica) infected with Pcrkinsiis marinus. Re.s. Rev. Parasit. 54:121-123. Cheng, T. C. & W. J. Dougherty. 1995. Partial inhibition of hemocyte agglutination by Lathyrus odoratus lectin in Crassostrea virginica in- fected with Perkinsiis marinus. Mem. Inst. Oswaldo Cruz. 90:407- 410. Cheng, T. C, J W. Huang, H. Karadognan, L R Renwrantz & T. P Yoshino. 1980. Separation of oyster hemocytes by density gradient centrifugation and identification of their surface receptors. J Imcri. Pathol. 36:35-40. Cheng, T. C.,W. J. Dougherty & V. G. Burrell, Jr. 1993. Lectin-binding differences on hemocytes of two geographic strains of the American oyster, Crassostrea virginica. Trans. Am. Microsc. Soc. 1 12:151-157, Cheng. T. C, W. J. Dougherty & V. G. Burrell, Jr. 1994. A possible hemocyte surface marker for resistance to Haplosporidium nelsoni in the oyster Crassostrea virginica. Res. Rev. Parasit. 54:51-54. Cheng. T. C. J. J Manzi & V. G. Burrell, Jr. 1995. Differences in lectin-binding by hemocytes of oysters (Crassostrea virginica) from three regions and further evidence for the correlation between the pres- ence of lathyrose and the absence oi Haplosporidium nelsoni. J. Shell- fish Res. 14:477-481. Dougherty, W. J.,T. C. Cheng &V. G. Burrell, Jr. 1993. Occurrences of the pathogen Haplosporidium nelsoni in oysters, Crassostrea virgi- nica. in South Carolina. USA. Trans. Am. Microsc. Soc. 112:75-77. Ford. S. E.. S. A. Kanaley, M. Ferns. & K. A. Ashton-Alcox. 1990. "Panning", a technique for ennchment of the oyster parasite Haplo- sporidium nelsoni (MSX). J. Invertbr. Palhol. 56:347-352. Gauthier. J. E. & W. S. Fisher. 1990. Hemolymph assay for diagnosis of Perkinsus marinus in oysters, Crassostrea virginica (Gmelin, 1871). J. Shellfish Res. 9:367-371. Kanaley, S. A. & S. E. Ford. 1990. Lectin binding characteristics of hemocytes and parasites in the oyster, Crassostrea virginica, infected with Haplosporidium nelsoni (MSX). Parasite Immunol. 12:633-646. Neter, J,, W. Wasserman & M, H. Kutner. 1983. Applied Linear Regres- sion Models. Richard D. Irwin. Homewood. Illinois. Ticha. M.. 1. Zemeddine & J. Kocourek. 1980. Studies on lectins XLVIII. Isolation and characterization of lectins from seeds of Lathyrus odo- ralus L. and Lathyrus silvestris L. Acta Biol. Med. Germ. 39:649-655. Yoshino, T. P., L. R. Renwrantz & T. C. Cheng. 1979. Binding and redistribution of surface membrane receptors of concanavalin A on oyster hemocytes. J. E.xp. Zool. 207:439^49. Journal of Shellfish Research. Vol, 15. No. 2. 395-400, 1996. PREVALENCE, INTENSITY, AND DETECTION OF BON AMI A OSTREAE IN OSTREA EDULIS L. IN THE DAMARISCOTTA RIVER AREA, MAINE ADRIANA I. ZABALETA AND BRUCE J. BARBER' Department of Animal. Veterinary & Aquatic Sciences University of Maine 5735 Hitclmer Hall Orono. Maine 04469 ABSTRACT Oysters, Osirea ediilis. collected from three locations in the Damariscotta River area, Maine, were examined for the presence ofBymjmiaoi/rcoc four times between June 1994 and April 1995. Overall prevalence was 5% (13 of 291), with no significant differences between sampling sites or dates. All but one of the infections were classified as "low" intensity. Results obtained with histological preparations and stained blood smears were similar. Results obtained with the immunofluorescence technique were unclear. No fluorescent B ostreae cells were observed with monoclonal antibody (MAB) 20B2. Even though MAS 15C2 was in 75% agreement with results obtained by standard histological preparations, (he fluorescence was faint. Condition index was not significantly different between infected and uninfected oysters. Although B ostreae is present in the Damariscotta River area, the development of disease and mortality may be precluded by the low density of oysters in natural beds and the relatively cold winters. KEY WORDS: Oysters, Ostrea eilnlis. Bonamia ostreae. condition index, disease INTRODUCTION Bonamiasis. a disease affecting oysters {Ostrea spp.), is caused by the protozoan parasite Bonamia ostreae (Ascetospora) iPichot et al. 1980). This parasite affects the flat oyster, Ostrea ediilis L., in several countries including France, Spain, the Netherlands, Ire- land, the United Kingdom, and the United States (Balouet et al. 1983, Elston et al. 1987. Monies and Melendez 1987. Farley et al. 1988. Friedman et al. 1989, McArdle et al. 1991. van Banning 1991. Friedman and Perkins 1994). Mortalities related to bonami- asis can be higher than 80% (Balouet et al. 1983, Monies and Melendez 1987. McArdle et al. 1991). Bonamiasis is character- ized by hemocytie infiltration around the stomach, style sac, and intestine, and the parasite is observed as small (2-3 \xm) "micro- cells" within hemocytes (Balouet et al. 1983, Elston et al. 1986, Sindermann 1990). In advanced cases, parasites are also seen in ciliated epithelial cells of the gills (Monies et al. 1994). The par- asite is phagocytosed by hemocytes (primarily granulocytes) and localized within a parasitophorous vacuole, where it multiplies and spreads to other tissues via the hemolymph (Balouet et al. 1983, Chagot et al. 1992, Monies et al. 1994). Transmission of the disease is likely related to the proximity of infected individuals, local currents, density of oysters, and the oyster's susceptibility to disease (Hudson and Hill 1991). Infected material (e.g., mud and organisms) moved from infected places to areas free of B. ostreae is another means by which the disease is spread (van Banning 1991). Susceptibility to bonamiasis varies between sites, age classes, and even between different individuals (Gnzel et al. 1988. van Banning 1991, Hervio et al. 1995). Fac- tors such as harvesting, replanting, storage, handling, and fluctu- ation in environmental parameters can also affect the occuirence of the parasite (Grizel et al. 1987, van Banning 1991, Caceres- Martinez et al. 1995). The diagnosis of B. ostreae is based on time-consuming his- topathological techniques. Two alternative methods of diagnosis that promise to be faster, less expensive, and more accurate have 'Corresponding author. been developed; stained heart smears and indirect antibody immu- noassay (Mialhe et al. 1988b, Boulo and Mialhe 1989). B. ostreae has recently been detected in O. edidis from the New Meadows River, Quahog Bay. and the Damariscotta River, ME (Barber and Davis 1994, Friedman and Perkins 1994). The Damariscotta River is the center of oyster culture in Maine because of favorable environmental conditions, including high primary productivity (Hidu et al. 1981). Oyster farmers are interested in culturing O. ediilis but are concerned about the potential effect of B. ostreae. Knowledge of the distribution of B. ostreae in the Damariscotta River area and its prevalence at various sites and times of year provides an important starting point for making decisions regarding the culture of flat oysters in Maine. The ob- jectives of this study were to determine 1 ) the seasonal prevalence and intensity of infection of B. ostreae in natural oyster beds in the Damariscotta River area, and 2) whether the condition index of oysters is affected by infection. Three diagnostic techniques for detecting B. ostreae (histological preparations, blood smears, and indirect immunofluorescence) were compared. MATERIALS AND METHODS Oysters were collected from three locations (each more than 6 km apart) within the Damariscotta River watershed; Little Point (Lat. 44°0rN; Lon. 69°32'W): Mears Cove (Lat. 43°57'N; Lon. 69°34'W); and Witch Island, Johns Bay (Lat. 43°52'N; Lon. 69°33'W) (Fig. 1). The density of oysters at each site was esti- mated by counting the oysters present in 10 ( l-m") quadrats along a transect. Water temperature was registered throughout the year with temperature-recording devices (Ryan RTM 2000 Thermo- graphs) installed at Little Point and at the Darling Marine Center (near the Mears Cove location). Records of temperature from the Maine Department of Marine Resources, Boothbay Harbor, were considered an approximation of temperature at Witch Island. At all stations, monthly average surface temperature was calculated from records obtained every 6 h. Scuba divers hand collected samples of 25 adult oysters of similar size (>70 mm in length) from each location on June 10th. August 25th, and November 30th, 1994, and on April 18th, 1995. Fouling organisms were removed before whole weights of each 395 396 Zabaleta and Barber Figure 1. Detail map of the Damariscotta River area, Maine, showing locations of sampling sites: Little Point, Mears Cove, and Witch Is- land. oyster were obtained. Oysters were then opened, and a cross- section (4-5 mm) adjacent to the labial palps containing the di- gestive gland, gonad, and gill tissues was removed and weighed (wet) before fixation for histopathologieal examination. The wet weight of the remaining tissue was recorded. The remaining tissue was then placed m an individual weighing dish, dried in a drying oven (82°C) to a stable weight, and held at room temperature m a desiccator before dry weight was recorded. Dry shell weight was recorded in a similar fashion. A dry weight:wet weight ratio of the remaining tissue was used to estimate the dry weight of the cross- section; the dry weight of the cross-section and the dry weight of the remaining tissue were added to obtain the dry weight of the entire animal (Barber et al. 1988). Condition index was determined for each individual as: dry soft tissue weight (g) x 1 ,000/intemal shell cavity capacity (g). where internal shell cavity capacity = the total whole live weight (g). minus the dry shell weight (g) (Hawkins et al. 1987). The condi- tion indices of uninfected oysters were examined with respect to site and sampling date by the use of two-factor analysis of variance without replication (Zar 1984). A Student's Mest (Zar 1984) was used to compare condition indices of infected and uninfected oys- ters collected in November from Little Point and Witch Island, because at Mears Cove and on other sampling dates, the parasite was either absent or was observed at very low prevalence. Histological Preparations Tissue cross-sections were fixed in Helly's fixative (June 1994 only) or Dietrichs' fixative (all other samples) (Barszcz and Yevich 1975) and processed for routine paraffin histology. Depar- affinized tissue sections (5 \xm) were stained with Shandon"s in- stant hematoxylin and eosin-Y and coverslipped with Flo-texx mounting medium. Stained tissue sections were examined with a Nikon Labophot microscope (lOOx). and the presence of hemo- cytic infiltration was recorded. The number of B. oslreae cells observed in 10 min (phase contrast, l.OOOx) in a randomly se- lected transect (including mantle, gonad, intestine, and digestive gland) of each histological preparation was recorded. The confi- dence intervals for Poisson random variables were calculated to detect differences in the prevalence of B. ostreae (number of in- fected oysters) among the three sites (Beyer 1968). Infection intensity was established by the procedure of Rogan et al. (1991). Oysters were ■'negative" if no parasites were ob- served within 10 min. In oysters with "'low" infections, fewer than 100 parasites were observed within 10 min. In "heavy" infections, practically all blood cells were parasitized. A Student's Mest (Zar 1984) was performed to compare the level of infection (logarithm of the number of parasites per infected oyster) of sam- ples collected in November from Little Point and Witch Island, the only samples where more than one parasite was detected. Blood Smears Cardiac tissue was used in the preparation of smears because it provides the most accurate diagnosis of B. oslreae infection (Boulo and Mialhc 1989). The heart (ventricle) of each oyster collected in June 1994 was removed, dried on blotting paper, pressed onto a slide, and air dried. Because of the excessive ag- gregation of hemocytes. oysters collected subsequently were pro- cessed as follows. A syringe (3 cm') was inserted into the peri- cardial cavity, and hemolymph (0.5 ml) was drawn into the barrel. The hemolymph was diluted with 2.5 ml of filtered (0.45-|j.m- pore-size filter) seawater. Aliquots (0.5 ml) of diluted hemolymph were placed on slides, and hemocytes were allowed to settle (15 min). Excess seawater was drained, and the smears were air dried. Smears were fixed (LeukoStat fixative solution. Fisher Diag- nostics), stained with Hemacolor blood stains (EM Diagnostic Systems), and mounted with Flo-texx mounting medium. The slides were observed for 10 min under the microscope (l.OOOx) and the number of parasites was recorded. Slides were considered positive only when the parasite was unambiguously observed. Slides were considered negative when, after two observations of more than 10 min each, either no parasite was detected or suspi- cious intrahemocytic inclusions (smaller diameter and without an evident nucleus) could not be positively identified as B. oslreae cells. McNemar's test (Zar 1984) was performed to compare the accuracy of histological preparations and stained blood smears in determining the prevalence of infection. A paired Student's (-test (Zar 1984) was used to compare the intensity of infection recorded by the two techniques. Indirect Immunofluorescence Air-dried smears were fixed in acetone for 8 mm and frozen at -20°C. The smears were overlaid with a monoclonal antibody solution (MAB). Two concentrations of MAB were tested. 50 and 100 [ig/ml (in phosphate buffer solution [PBS]: phosphate. 10 mM: NaCl, 150 niM; pH 7.4). After incubation (30 min) at room temperature in a moist chamber, the slides were washed with PBS and then overlaid w ilh fluorescein isothiocyanate-conjugated anti- B. osTREAE IN Maine 347 Tabk' 1. Kstimated Densitj of Oysters, O. edulis. Substrate Type, and Depth Range (Low Tide) at Each Collection Site. TABLE 2. Number of Oysters, O. edulis. Infected with B. ostrea (as Determined by Histological Examination), by Site and Sampling Date; Sample Size Indicated in Parentheses. Density (Oysters/m") Substrate Depth (m) Location Location 10 Jun 1994 25 Aug 94 30 Nov 1994 18 Apr 1995 Little Point Mears Cove 0.5 0.33 0.04 Shell hash, cobble, sandy-mud Soft mud Shell hash, sandy mud 2-5 5-8 2-4 Little Point Mears Cove Witch Island 0(25) 0(25) 0(25) 0(23) 0(25) 0(25) 8 (25) 0(25) 3(25) 1 (25) 1 (25) 0(17) Witch Island mouse immunoglobulin antiserum (Sigma Immunoehemieals) di- luted in PBS containing 0.01% Evans-Blue. The slides were again incubated and waslied as described. The slides were mounted with glycerin buffer solution and immediately examined for fluores- cence with a Nikon Fluophot microscope at l,000x (Boulo and Mialhe 1989). B. oslreae cells were counted over a 10-min period. Two B. oslreae MAB, 20B2-1B12 and 15C2-2F2 (IFREMER. Montpellier. France), were tested. The technique was tested with oysters collected in November from the three locations and with oysters collected in April from Little Point and Mears Cove. Re- sults from the histopathological technique were used as the con- trol. RESULTS The Little Point and Witch Island sites had similar bottom characteristics, being primarily composed of shell hash and sandy- mud substrate (Table I ). Both sites were fairly shallow, with depth ranging from 2 to 5 m at low tide. In contrast. Mears Cove had a softer substrate (mud) and was deeper at low tide. 5-8 m. The highest estimated density of oysters (0.5/nr) was recorded at Little Point (Table I). The monthly average temperature at Little Point was higher than that at the other two sites from May to September, ranging from 13 to 22.5°C (Fig. 2). At Mears Cove and Witch Island, the range for the same period was 9-18°C. In total, 291 oysters were processed throughout the year of sampling. Hemocytic infiltration was evident in 16% (45 of 291) of the oysters. Only 5% (13 of 291 ) of all the oysters examined throughout the year were infected with B. ostreae. Suspicious intrahemocytic inclusions were seen in another 5% { 13 of 291). Of the 13 infected oysters, 31% (4 of 13) had focal inflammation of connective tissue, and B. ostreae cells were found only in those 25 o ^u o 15 3 ra k. Q) 10 Q. E (1> 1- b ■«■' -■•-- Little Pt — *— Mears C -■■■- Witch 1 - --^,.-'- Jan March IVIay July Sept Nov Jan Marcti Month Figure 2. Monthly average water temperatures from the three sample locations (Little Point, Mears Cove, and Witch Island) from January 1994 to April 1995. foci. Concurrent inflammation and B. ostreae infection were seen in 47% (9 of 19). 7%^ ( 1 of 13). and 14% (2 of 14) of oysters from Little Point. Mears Cove, and Witch Island, respectively. The parasite was obsei^ed in gill tissue in most of the infected oysters, even when not detected in connective tissue around the intestine or digestive gland, and was present in the intestinal wall only in more highly parasitized oysters. Prevalence and Intensity of B. ostreae In this study, even though approximately 8.7% ( 13 of 148) of oysters collected between June and August had suspicious intra- hemocytic inclusions, typical B. ostreae cells were not observed before November 1994 (Table 2). The confidence interval for Poisson random variable indicated that there were no significant differences (P > 0.05) among sites in the prevalence of B. ostreae in November 1994 (confidence limits; Little Point, 3.4-15.8; Mears Cove, 0.0-3.7; Witch Island, 0.6-8.8) and April 1995 (Lit- tle Point. 0.0-0.2; Mears Cove. 0.0-0.2; Witch Island. 0.0-0.2); there were no significant differences (P > 0.05) between Little Point (0.1-0.6) and Witch Island (0.0-0.2) for samples collected in November 1994 and April 1995. The intensity of B. ostreae infection, on the basis of the ex- amination of histological preparations, was "low" in most of the infected oysters (Table 3). Only one individual (Little Point. No- vember 1994) had a "heavy" infection (practically all blood cells parasitized). The intensity of B. ostreae infection, however, was not significantly different between Little Point and Witch Island oysters collected in November (f-test. P = 0.46). Condition Index Two-factor analysis of variance without replication revealed no significant differences in condition index (range, 82.2-217.6) as- sociated with either site (P = 0.28) or sampling date (P = 0.20) (Table 4). Even though infected oysters collected in November TABLE 3. Intensity of B. oslreae Infection as Determined From Histological Preparations of O. edulis. 10 Jun 1994 25 Aug 1994 30 Nov 1994 18 Apr 1995 Location M R M R M R M R Little Point Mears Cove Witch Island 0 — 0 — 0 — 0 0 0 — 9.5 0 10 4-334 1-20 1 — 1 — 0 — M. median values of number of parasites seen in 10 min in each infected oyster, by site and sampling date; R. range of parasite number. 398 Zabaleta and Barber TABLE 4. Gravimetric Condition Indices of Oysters, O. edulis. Uninfected and Infected Witli B. ostreae, by Site and Sampling Date. Little Point Mears C ove Witch Island Uninfected Infected Uninfected Infected Uninfected Infected Date M SD M SD M SD M SD M SD M SD 10 Jun 1994 25 Aug 1994 30 Nov 1994 18 Apr 1995 107.0 89.9 120.7 113.9 19.8 28.0 45.4 37.4 106.9 45.7 67.2 ~ 138.0 82.2 217.6 133,4 38.5 23.7 527 33.9 52.1 — 115.3 136.2 139.7 114,7 35.0 106 41.8 25.8 97.4 27.2 M, mean; SD, standard deviation. 1994 at Little Point and Witch Island tended to have lower average condition index (104.4 ± 40.4) than noninfected oysters (131.2 ± 44.5), the difference was not significant (/-test; P = 0.892). Comparison of Techniques The quahty of blood smears prepared from heart tissue col- lected in June 1994 was unsuitable for parasite screening because of dense cell aggregations. Subsequent smears prepared from di- luted blood were satisfactory. Even though fi. ostreae cells stained with Hemacolor were easily observed, diagnosis was uncertain in 2% (5 of 217) of oysters. The parasite was observed principally in agranular cells. Although 13 oysters were determined by histological prepara- tion to be infected, only 6 were detected with stained blood smears. In contrast, three infected oysters, detected with blood smears, were considered uninfected after histological examina- tion. In each of them, however, only one parasite was observed. The McNemars's test, used to compare the prevalence of mfection detected by histological preparations and stained blood smears, revealed no significant differences between the two techniques (Z = 0.90). In five of the six infected oysters detected by both techniques, histological preparations detected a 50% greater num- ber of parasites than did stained blood smears. A comparison of the mean differences between the number of parasites detected by histology and blood smears, however, showed that the techniques were not significantly different (P = 0.34). Low-intensity infec- tions characterized by focal hemocytic infiltration were detected only by histological examination. Histological results were used as references for immunological tests performed on samples collected in November and April. Sec- tions of uninfected animals, used as controls, showed no back- ground or nonspecific fluoresence with both 20B2-IB12 and 15C2-2F2 at both concentrations (50 and 100 |ig/ml PBS), fi. oi/reae'-positive slides tested with 20B2 did not fluoresce at either 50 or 100 M-g/ml PBS. Seventy-three percent of B. ostreae- positive slides tested with I5C2-2F2 showed faint fluorescence associated with parasite cells. Results from one slide were inde- terminate, and results from another were negative. The fluorescence pattern was similar at both MAB concentrations (50 and 100 ^.g/ml). DISCUSSION fi. ostreae was found at all three sampling sites; however, no obvious B. ostreae cells were detected before November 1994 at any site. At Mears Cove and Witch Island, prevalence was 8% or less throughout the year of sampling. Prevalence was 32% at Little Point in November 1994. Past prevalences of B. ostreae infections at Little Point, detected by the use of histological preparations, were 34% in June 1991. 45% in June 1992. and 20% in August 1993 (Barber and Davis 1994, Friedman and Perkins 1994). Thus, the prevalence of B. ostreae in the Damariscotta River at Little Point has been relatively steady since the parasite was first de- tected in 1991. The finding of low prevalence and low infection intensity in- dicated a general lack of disease development in oysters parasit- ized by B. ostreae in Maine. In addition, a high percentage of oysters had focal hemocytic infiltration, rather than general in- flammation. Ninety-two percent of infected oysters had a '"low" infection intensity: in many, fewer than 10 parasites were ob- served. Even though infected oysters examined in 1993 had a higher intensity than those in this study, intensity was still within the "low"" range of infection (B. Barber unpublished data). The prevailing low prevalence and low infection intensity of B. ostreae infection in Maine oysters may be related to low densities of oysters and prevailing climatic conditions in Maine (long, cold winters followed by short, mild summers), both of which restrict the spread of the disease (Elston and Holsinger 1988). Accord- ingly, the highest prevalence and intensity of B. ostreae recorded in this study occurred at Little Point, where oyster density and water temperature were the greatest. Parasitic infections have been shown to reduce the condition index of bivalves (Montes and Melendez 1987. Barber et al. 1988). The results of this study revealed no significant differences between condition indices of infected and uninfected oysters col- lected in November at Little Point and Witch Island, when infec- tion intensity was greatest. Rogan et al. (1991) suggested that condition indices were not appreciably affected in oysters with low infection intensity. All but one infected oyster found in this study had low intensities of infection; this may have contributed to the lack of detection of a significant difference in condition index between infected and uninfected oysters. This lack of effect on condition index is another indication that levels of B. ostreae in the Damariscotta River in 1994-1995 were generally not great enough to cause disease. The detection of B. ostreae is difficult in animals with a low infection intensity. The absence of infected oysters in samples collected between June and August 1994 could be the result of the misdiagnosis of lightly infected oysters, a small sample size, or a combination of both factors. A larger sample size would be nec- essary to detect B. ostreae in areas with both low prevalence and low infection intensity (Ossiander and Wedemeyer 1973). Unfor- tunately, a scarcity of oysters limited sample size throughout this study. Histological preparations detected more infected oysters than B. OSTRhAE IN MaINE 399 did stained blood smears, although the difference was not signif- icant. Failure to detect the parasite with stained smears at low infection intensity has been previously reported (Bucke 1988, Bucke and Feist 1985. McArdle el al. 1991). Histological prepa- rations are a time-consuming and expensive diagnostic technique. Without significant differences in detection between these tech- niques, stained blood smears are thus adequate for preliminary studies or in systematic screening to help determine infection sta- tus in a particular area. By the use of blood smears, a larger sample can be screened rapidly, without sacrificing oysters. Histological preparations should be reserved for more exhaustive research, or for situations where the detection of light (early) cases is important (e.g., before moving oysters to areas free of 5. oslreae). The immunofluorescence technique developed by Miaihe et al. (1988b), tested extensively for the first time on oysters from Maine, gave unclear results. MAB 20B2 did not bind B. ostreae cells on any of the tested slides. Fluorescence of S. ostreae cells was observed in 73% of the positive slides tested with I5C2. but the pattern was faint. The differential performance of the MAB solutions may be explained by a different decline in the affinity for an epitope caused by an excess of seawater (V. Boulo pers. comm.). No antigenic differences have been observed between B. ostreae from Washington and from Europe (Miaihe et al. 1988a). Potential serological differences between B. ostreae and the par- asite observed in this study, however, cannot be ignored. B. os- treae has been observed in all hemocyte types in O. edulis but is found primarily in granulocytes (Balouet et al. 1983, Mourton et al. 1992). The fact that the parasite in this study was observed principally in agranular cells might also mdicate a difference be- tween B. ostreae from Maine and from Europe. Additional im- munofluoresence assays should be developed to improve this tech- nique, which has been shown elsewhere to be as effective as histology and stained blood smears in detecting the disease (Boulo and Miaihe 1989). In addition, because the immunofluorescence assay is an important taxonomic tool (Miaihe et al. 1988a), it could be used to ultimately verify the identity of B. ostreae in Maine. B. ostreae is currently resident in several locations in Maine (Barber and Davis 1994, Friedman and Perkins 1994, this study). Even though the prevalence and intensity of the parasite are at present relatively low, the potential for oyster mortality caused by bonamiasis remains. Further experiments growing oysters under commercial conditions (high density) will be necessary to deter- mine the potential effects of B. ostreae on the commercial culture of O. edulis in Maine. ACKNOWLEDGMENTS This work was supported by a Fulbrighl-LASPAU scholarship to A.l.Z. Thanks to Chris Davis, Alvaro Palma, and Laurie Steams for collecting oysters. Dr. W. Halteman helped with the statistical analysis, and Drs. D. Hervio and C. Moody provided advice on the immunofluorescent technique. Thanks to Dawna Beane for assisting with histological preparations and J. P. Cado- ret and V. Boulo (IFREMER) for providing the antibody solu- tions. Dr. R. Hillman and Mr. A. Farley screened several samples, and the Maine Department of Marine Resources and Damariscotta River Association provided temperature and salinity data. Support of the Maine Aquaculturc Association and the Maine Aquaculture Innovation Center is gratefully acknowledged. This is Maine Ag- ricultural and Forest Experiment Station external publication #1979. LITERATURE CITED Halouet. G.. M. Poder & A. Cahour. 1983, Haemocylic parasitosis; mor- phology and pathology of lesions in the French flat oyster, Oslrea edulis L. Aquaculture 34:1-4. Barber, B. J. & C. Davis. 1994. Prevalence of Bonamia oslreae in Ostrea edulis populations in Maine. J. Sliellfisli Res. 13:298. Barber. B J . S. E. Ford & H. H. Haskin. 1988. Effects of the parasite MSX {Haplosporidian nelsoni) on oyster (Crassostrea viginica) energy metabolism. I. Condition index and relative fecundity. J. Shellfish Res. 7:25-31. Barszcz, C. A. & P. P. Yevich. 1975. The use of Helly's fixative for marine invertebrate histopathology. Comp. Pathol. Bull. 7:4. Beyer, W. H, 1968. CRC Handbook of Tables for Probability and Statis- tics. Chemical Rubber Co., Cleveland. Ohio. 642 pp. Boulo, V. & E. Miaihe. 1989. Immunodiagnosis of Bonamia oslreae (Ascetospora) infection of Ostrea edulis L. and subcellular identifica- tion of epitopes by monoclonal antibodies. J. Fish Dis. 12:257-262. Bucke, D. 1988. Pathology of bonamiasis. Parasitol. Today 4(6):174- 176. Bucke. D. & S. Feist. 1985. Bonamiasis in the tlat oyster. Oslrea edulis. with comments on histological techniques, pp. 387-392. In: A. E. Ellis (ed.). Fish and Shellfish Pathology. Academic Press, London, England. Caceres-Marti'nez. J., J. A. F. Robledo & A. Figueras. 1995. Presence of bonamia and its relation to age, growth rates and gonadal development of the flat oyster, Oslrea edulis. in the Ria de Vigo. Galicia (NW Spain). Aquaculture 130:15-23. Chagot. D., V. Boulo, D. Hervio. E, Miaihe. C. Mourton & H. Gnzel. 1992. Interactions between Bonamia ostreae (Protozoa: Ascetospora) and hemocytes of Ostrea edulis and Crassostrea gigas (Mollusca;Bi- valvia); entry mechanisms. J. Inverlebr. Pathol. 59:241-249. Elston, R. & L. Holsinger. 1988, Resistant flat oysters offer hope against bonamiasis. Parasitol. Today 4(5):120-121. Elston. R. A.. C. A. Farley & M. L. Kent. 1986. Occurrence and signif- icance of bonamiasis in European flat oysters Oslrea edulis in North America. Dis. Aqual. Org. 2:49-54. Elston, R. A., M. L. Kent & M T. Wilkinson, 1987. Resistance of Os- lrea edulis to Bonamia ostreae infection. Aquaculture 64:237-242. Farley, C. A., P. H. Wolf & R. A. Elston. 1988. A long-tenn study of "microcell" disease in oysters with a description of a new genus. Mikrocylos mackini (sp.n) and Mikrocylos roughleyi (sp.n. ). Fish Bull 86:581-593. Friedman. C. S. & F. O. Perkins. 1994. Range extension of Bonamia ostreae to Maine, U.S. A, J. Inverlebr. Pathol. 64:179-181. Fnedman, C. S., T. McDowell, J. M. Groff, J. T. Hollibaugh, D. Man- zer & R. P. Hedrick. 1989. Presence oi Bonamia ostreae among pop- ulations of the European flat oyster. Oslrea edulis Linne, in California. U.S.A. J. Shellfish Res. 8:133-137. Gnzel. H.. E. Bachere. E. Miaihe & G. Tige. 1987. Solving parasite- related problems in cultured mollusks, pp. 301-308. In: M. J. Howell (ed.) Proceedings of the Sixth International Congress of Parasitology. Australian Academy of Science, Brisbane. Grizel, H., E. Miaihe, D. Chagot, V. Boulo & E. Bachere. 1988. Major parasitic and pathological conditions. Bonamiasis; a model study of diseases in marine mollusks. Am. Fish. Soc. Spec. Publ. 18:1-^ Hawkins, C. M., T. W. Rowell & P. Woo, 1987. The importance of cleansing in the calculation of condition index in the soft-shell clam. Mya arenaria (L). J. Shellfish Res. 6:29-36. Hervio, D.. E. Bachere, V. Boulo. N, Cochennec. V, Vuillemin, Y, Le Coguic. G. Gailletaux. J. Mazurie & E. Miaihe. 1995. Establishment of an experimental infection protocol for the flat oyster. Oslrea edulis, 400 Zabaleta and Barber with the intrahemocytic protozoan parasite. Bonamia ostreae: applica- tion in the selection of parasite-resistant oysters, Aquaculmre 132: 183- 194. Hidu, H., S. R. Chapman & D. Dean. 1981. Oyster mariculture in suh- boreal (Maine, United States of America) waters; cultchless setting and nursery culture of European and American oysters. J. Shellfish Res. 1:57-67. Hudson, E. B. & B. J. Hill. 1991. Impact and spread of bonamiasis in the U.K. Aquaculmre 93:279-285. McArdle, J. F., F. McKieman, H. Foley & D, H. Jones. 1991. The cur- rent status of Bonamui disease in Ireland Acjucuulture 93:273-278. Mialhe, E., V. Boulo, R. Elston. B. Hill, M. Hine. J. Monies. P. Van Banning & H. Gnzel. 1988a. Serological analysis oi Bonamia in Os- trea edulis and Tiostrea lutharia using polyclonal and monoclonal antibodies. Aqual. Living Resour. 1:67-69. Mialhe, E., E. Bachere. D. Chagot & H. Grizel. 1988b. Isolation and Purification of the protozoan Bonamia ostreae (Pichot et al. 1980), a parasite affecting the flat oyster Ostrea edulis L. Aquacuhure 71:293- 299. Montes, J. & M. I. Melendez. 1987. Donnes sur la parasitose de Bonamia ostreae chez I'huitre plate de Galicie. cote nord-ouest de I'espagne. Aquacullure 61 •.\95-l9». Montes. J., R. Anadon & C. Azevedo. 1994. A possible life cycle for Bonamia ostreae on the basis of electron microscopy studies. J. In- veriebr. Pathol. 63:1-6. Mourton, C, V. Boulo, D. Chagot. Hervio, E. Bachere. E. Mialhe & H. Gnzel. 1992. Interactions between Bonamia ostreae (Protozoa: Asc- etospora) and hemocytes of Ostrea edulis and Crassostrea gigas (Mol- lusca: Bivalvia): in vitro system establishment. J. Inveriebr. Pathol. 59:235-240. Ossiander, F. J. & G. Wedemeyer. 1973. Computer program for sample sizes required to determine disease incidence in fish populations. J. Fish. Res. Board Can. 30:1383-1384. Pichot, Y., M. Comps, G. Tigue, H. Grizel & M. A. Rabouin. 1980. Recherches sur Bonamia ostreae gen. N., sp. n.. parasite nouveau de I'huitre plate Ostrea edulis L. Rev. Trav. Inst. Peches Marit. 43:131- 140. Rogan. E.. S. C. Culloty. T, F. Cross & M. F. Mulcahy. 1991. The detection of Bonamia ostreae (Pichot et al. 1980) in frozen oysters (.Ostrea edulis L.) and the effect of the parasite on condition. Aqua- culture 91 .i\\-M5 . Sindermann. C. J. 1990. Pnncipal Diseases of Manne Fish and Shellfish, vol. 2. Academic Press. San Diego. California. 516 pp. van Banning. P. 1991. Observations on bonamiasis in the stock of the European flat oyster. Ostrea edulis. in the Netherlands, with special reference to the recent developments in Lake Grevelingen. Aquaculture 93:205-211. Zar. J. H. 1984. Biostatistical Analysis. Prentice-Hall. Inc., Englewood Cliffs. New Jersey. 718 pp. Journal of Shellfish Research. Vol. 15, No, 2. 401^06. 1996. FACTORS AFFECTING THE GRAZING RATE OF THE NEW ZEALAND ABALONE HALIOTIS IRIS MARTYN ISLAY D. MARSDEN AND PAUL M. J. WILLIAMS Zoology Deparlmeni University of Canterbury Chrisichunh, New ZeuUmd ABSTRACT Grazing rates of the New Zealand black foot paua Halioris iris Martyn were measured for Individuals supplied with four species of shallow subtidal algal food. Measurements were made on II length groups of paua. 16-120 mm. Feeding rates depended on body size, with a decrease in the weight-specific grazing rate with increasing shell length up to 66 mm. Grazing rates for rhodophytes, Gracilaria sp. and Hymenocladia sanguinea exceeded those on phaeophytes. Macrocystis pyrifera and Lessonia variegata. The assimilation efficiency of paua feeding on Gracilaria sp. was the highest of the four species tested (85%). During winter, abalone grazing rates at a temperature of 10°C were significantly reduced compared with summer values at 17.5°C. Con- sumption rates on the kelp L. variegata were similar during summer and winter. Laboratory experiments investigated the effects of density on the feeding and growth rate of juvenile paua, initial length 35-45 mm, over a year. Although feeding rates were relatively constant with increasing density, annual length increment decreased. At densities greater than 40 individuals per cubicle (400 m~"). there was no length increase and individuals lost weight. The highest growth rate was found at the lowest density, where an annual length increment of 9.2 mm was about half that recorded from field populations. The grazing rates of the black foot paua feeding on natural algal food are discussed in relation to their natural diet, and the algae are evaluated for their potential use as a food supply for abalone manculture. KEY WORDS: Abalone. grazing rates, growth, gastropod. Haliolis iris INTRODUCTION Haliotis iris, the common or black foot paua, is the largest of the three species of abalone found in New Zealand. It is discon- tinuously distributed along the coast of both main islands in New Zealand. Stewart Island and Chatham Islands (Poore 1969). Al- though there has been worldwide interest in the ecology and growth of abalone in different parts of the world (Shepherd 1973, McShaneetal. 1988, Tegner et al. 1989, Day and Fleming 1992), there is relatively little information about the New Zealand spe- cies, which are of increasing interest for commercial culture. Eco- logical studies on H. iris have provided details of natural diet, and growth rates have been calculated from field studies (Poore 1972a. Poore 1972b, Sainsbury 1982, Pirker 1992). These studies suggest variable growth rates depending on geographical location, sea- sonal temperatures, and food supply. Recent studies on the feeding physiology of gastropods suggest that dietary composition depends on many factors, including food quality, quantity, palatability. and nutritional value (Paine and Vadas 1969, Steneck and Watling 1982. Carefoot 1987. Brendel- berger 1995). Food choice may not be predictable, with some species of gastropod herbivores showing food preferences based on the physical structure of the food rather than its nutritional value (Andrew 1986. Chew 1984). Food availability may also determine the dietary composition of some gastropods including abalone (Shepherd and Steinberg 1992). Poore (1969) examined the gut contents of H. iris from subtidal areas close to our labo- ratory and found that they fed mainly on red algae throughout the year. In contrast, paua collected 200 km south in a more protected locality fed inainly on drift weed consisting mainly of the brown kelp Macrocystis pyrifera. This study was prompted by recent interest in the culture of the black foot paua using natural food supplies. Our goals were to estimate assimilation efficiencies and laboratory grazing rates of paua supplied with four different species of macroalgae. Three of these were locally abundant and known to be part of the natural diet of the paua (Poore 1972a, Sainsbury 1977). The fourth spe- cies, Gelidium species, was included because it has been identified as a potential food source for the commercial culture of abalone elsewhere (Pickering 1986, Tong 1986). Laboratory experiments were designed to investigate the effects of body size and season on the feeding rate of abalone. In addition, because of interest in land-based paua farming or barrel culture, the effects of density on algal consumption and the growth of//, iris were investigated by a laboratory experiment extending over a period of a year. MATERIALS AND METHODS A preliminary feeding experiment was undertaken to determine an appropriate level of food supply for the paua and a duration for the consumption experiments. This would be the time period at which variance of the mean consumption rates were least. Five individuals from three length groups (36-40, 66-70, and 96-100 mm) were supplied with a known wet weight of A/, pyrifera. one of the common species of drift algae found at Kaikoura. The paua were maintained in aquarium conditions with running seawater at a temperature of I2°C. The amount of W. pyrifera consumed by each individual was recorded every 24 h over a 15-d period. Daily consumption rates for each size group were calculated as dry weight of algal frond consumed per individual. This experiment showed that, for all three size groups, the variance in the daily algal consumption decreased from 1 to 5 d and then remained similar. As expected, small //. iris consumed more algae per dry weight of tissue than larger individuals. On the basis of these results, we chose to investigate the feeding rates of paua using a standard 5-d trial period. The grazing rates of //. iris were estimated during winter (Au- gust) when the aquarium seawater temperature was IO°C and sum- mer (January) at a temperature of I7.5°C. These temperatures were approximately 1 .5-2°C above the coastal surface seawater temperatures. The paua were collected from intertidal and shallow subtidal areas close to the marine laboratory. They were separated 401 402 Marsden and Williams into 11 5-mm-length groups ranging from 16 to 20 mm (first-year juveniles) to the largest size group, 1 16 to 120 mm in length. Any paua that had been damaged during collection (broken shell or damaged foot) were excluded. Experimental animals were allowed to acclimate to laboratory conditions over a period of 2 wk. during which they were supplied with a mixed selection of the algae that would be presented to them during the experiments. Faeces and any other material that accumulated at the bottom of the flow- through tanks were removed daily. The tanks received a low-level natural light regime. Four species of macroalgae were used in the experiments — two phaeophytes, M. pyrifera and Lessonia variegata. and two rhodophytes. Hxmenocladia saiiguinea and Gracilaria species. The Gracilaria sp. was collected weekly from intertidal areas close to Christchurch and transported to Kaikoura, where it was kept at 4°C. The other three species were freshly harvested from the same site every 2 d. At the start of each 5-d experiment, the length (in millimeters) and wet weight of five individuals from the 1 1 size groups were measured. Individuals were held separately and supplied with a known wet weight of one of the test algae. Algal treatments in each holding tank were assigned randomly. After each 24 h, the remaining algae were removed, reweighed, and replaced with a known quantity of fresh algae. Control tanks containing algae but without animals were used to test for any deterioration or decom- position of the algal food source. Algal wet weight loss was con- verted to dry weight loss by comparing wet/dry weight relation- ships from subsamples of algae taken at regular intervals and oven dried at 60°C. The dry weight of tissue from individual paua was estimated from a single sample (n = 30) covering the whole length range. Daily algal consumption rates were calculated for each size group, and consumption rates during winter and summer were compared by use of analysis of variance (ANOVA) (Snede- cor and Cochran 1976). If the variances were not homogeneous (Bartlett's test), trends were determined by the use of the nonpara- metric Kruskal-Wallis test (Siegel 1956. Sokal and Rohlf 1981). The Scheffe comparison of means test was used to compare means of the different size groups. The assimilation efficiencies of H. iris feeding on the four freshly collected test algae were determined by standard bomb calorimetry (Phillipson 1964). Eight paua from each of three length groups were used; 36-40, 66-70, and 96-100 mm in shell length. Animals were acclimated for 2 wk, during which they were supplied with a mixture of the test algae. They were starved for 72 h to evacuate their gut contents before being presented with a preweighed sample of a single alga. After 48 h, the remaining alga was removed and reweighed to estimate the feeding rate. Faeces from the group were collected for up to 48 h. These were washed quickly in distilled water to remove salts and oven dried at 60°C. Other groups of paua were held in similar conditions without food as control groups. The energy value of each algal species was calculated for freshly collected specimens and for those held in a circulating seawater system for 5 d. This latter sample was used to simulate drift algae and to evaluate its potential value as a food source for paua. The assimilation efficiency for each size group of paua was calculated by difference in the energy content (Kcal ■ g~' dry weight) between ingested and egested food ma- terial. The effects of density on the feeding and growth rates of ju- venile paua (length, 35^5 mm) were evaluated for groups of 2, 10, 20, 40, 60, and 80 individuals in each cubicle (39 x 27.5 x CD IS: 'k WINTER a Grocilorio sp. • Hymenoclgdio sonQuineg < Lessonia voriegoto o MocrocYStis pvriferg Figure during Size Class (mm) 1. Effect of body length on the algal consumption of H. iris winter. Values are mean values ±1SE; n = 25 for each species. 8 cm depth; surface area. 2,136 cm"). Polythene tubes were cut in half and placed in the cubicles to provide shelter. Preweighed algal food was supplied to the groups at 24-h intervals when any re- maining alga was removed. There were four replicates for each density. Individual paua were measured initially after the first and second month and then at 2-mo intervals for a year. Abalone grazing rates were measured during autumn (March), as described in the previous section. RESULTS Effects of Body Size and Season on Algal Consumption During winter, the mean daily consumption rates for H. iris varied with both size and food type (Fig. 1); smaller abalone showed increased consumption compared with larger individuals. The Scheffe comparison of means test showed similar consump- tion rates for adjacent length groups, and for individuals above 56 mm, the daily consumption rates were similar. Summer mean consumption rates of abalone were higher than winter values, reaching 18.7% of the body weight for the 16-20 mm-length group (Fig. 2). Mean consumption rates of adjacent length groups were similar (Scheffe comparison of means test), and three size groups, differentiated. These were smaller individ- '^, 10-' SUMMER i\ \ k. Grqcilqrig sp. Hymenoclodio sonquinao Lessoniq varieqato Size Class (mm) Figure 2. Effect of body length on the algal consumption of H. iris during summer. Factors Affecting H. iris Grazing Rate 403 TABLE 1. ANOVA Comparing the Effects of Size, Season and Algal Food Type on the Grazing Rate of Paua. Season Variable df P< Winter Size class (Kruskal-Wallis) n = 220 110.8 0.001 Algal species 3.219 2.2 NS Rhodophyte/phaeophyte 1.219 124 0.001 Summer Size class 10,2 30.9 0.001 Algal species 3,219 25.2 0.001 Rhodophyte/phaeophyte 1,219 57.2 0.001 The data were normalised by log" transformation. NS indicates a value that is not significant. TABLE 3. Caloric Values (Kcal g ' Dry Weight) for Fresh and Experimentally Produced Drift Algae. Assimilation Energy Algae Fresh Drift % Change Efficiency Value Gracilaria sp. 4.4 4.25 -3.4 85.4 3.76 H sangiiinea 3.6 3.25 -9.7 65.3 2.35 L. variegata 3.8 3.75 -1.3 64.3 2.44 M. pyrifera 3.9 3.80 -2.6 75.0 2.93 Also shown are the % assimilation efficiencies of H. iris feeding on par- ticular algal species. Energy value is the potential Kcal ■ g ' algae avail- able lo H iris. uals (16-30 mm in length), intermediate individuals (36-60 mm), and larger individuals (66-120 mm). By the use of log'' trans- formed results, it was found that winter feeding rates differed significantly from summer values (Kruskal-Wallis F = 40.3; P = 0.001; n = 440). Analyses of these results grouped into size groups suggest that for the largest individuals there was little vari- ation in the algal consumption rates with season. Effect of Algal Species on Paua Feeding Rates In all experiments, individuals less than 70 mm in body length consumed more Gracilaria than Hymenocladia. with L variegata and Macrocystis being eaten in similarly lesser quantities. In sum- mer, abalone consumption of all four algae differed, whereas dur- ing winter, when overall consumption rates were reduced, the paua consumed similar amounts of algal species grouped into their taxonomic groups. The winter grazing rates of abalone on the red algae were significantly less than that recorded for the brown algae (ANOVA; Table 1). Results from all size groups were combined to compare the seasonal grazing rates on the four algal foods. These showed significantly greater summer consumption by the abalone for three of the four species tested (Table 2). The paua grazing rate on the small kelp L. variegata was similar in summer and winter. Assimilation Efficiencies The caloric values (Kcal • g~' dry weight) of fresh and exper- imentally produced drift seaweeds are shown in Table 3. Gracilaria species had the highest value of the four species tested, and values remained high in all species except Hymenocladia after 5 d of immersion. The assimilation efficiency of//, ins feeding on Gracilaria exceeded 85%. The three other algal species had a similar caloric value, but the assimilation efficiency of paua feed- ing on Macrocystis exceeded that for the two red algae. When calculated as the potential energy gain per unit dry weight of algal TABLE 2. ANOVA Comparing the Effect of Season (Winter/Summer) on the Grazing Rates of Paua Feeding on Four Different Algal Species. Algal Species df Gracilaria sp. 1,109 29.2 0.001 H. sanguinea 1,110 23.4 0.001 L. variegata 1,110 0.2 NS M. pyrifera 1.110 13.4 0.001 food, Gracilaria remained the most efficient food, followed by Macrocystis. Lessonia, and Hymenocladia. Density Experiments The effect of density on the algal consumption of juvenile paua held long term in aquarium conditions is shown in Figure 3. Feed- ing rates ranged from '&.l'7c of mean body weight per day at the lowest density to 5.5% at a density of 60 individuals in each container. One-way ANOVA indicated no significant effect of density on the feeding rate (F = 2.39; P = 0.08; df = 5, 23), but paired tests detected differences between feeding rates of paua held at densities of 2 and those held at 60 or 80 individuals in each container. During the density experiment, mortality was low. Of the nine individuals that died over the year, six deaths occurred in the first month. These were replaced by individuals of a similar shell length that had been held for a similar period in the laboratory. At the three lowest densities, length and weight increases were sea- sonal (Fig. 4), with the highest values from November to February and little growth between March and August. The bimonthly mean length increase did not exceed 3 mm. corresponding to a wet weight increase of approximately 2.5 g at the lowest density. Over the experimental period of a year, there was no significant length increase in paua maintained at densities of 40. 60. and 80 indi- viduals per container (Fig. 5). Individuals held at the two highest densities lost weight over the exposure period. The effects of density on the annual weight and length increase 20 40 Paua Density Figure 3. Effect of stock density on the consumption rate (mean ± SE) of H. iris feeding on M. pyrifera during autumn. 404 Marsden and Williams of H. iris fed M. pyrifera are shown in Figure 5. Growth rate declined with increasing stock density, and paua lost weight at densities of more than 40 individuals per container. The greatest growth rates occurred at the lowest density, where the 9.2-mm increase in body length represented a doubling of the initial wet weight. DISCUSSION Laboratory grazing rates of the black foot paua depend on body size, food type, season, and density. Although the feeding rates were maintained at relatively high levels, the slow growth rates achieved during this study might suggest relatively poor potential for mariculture using natural food sources. Body size is a major factor affecting the feeding rates of gas- tropods. As in previous studies, the weight-specific grazing rates of smaller paua exceeded those for larger individuals (Paul et al. 1976, Day and Fleming 1992). However, the grazing rate re- mained constant above 60 mm in length. In H. iris, sexual matu- rity occurs between 60 and 80 mm shell length (Poore 1972b) with gonad maturation over winter. The relative independence of the grazing rate beyond the size at sexual maturity most likely reflects changes in energy allocation. This adaptation would allow com- paratively less energy to be used in body maintenance and shell and somatic growth and more energy to develop reproductive tis- sues. Some support for this can be found from fecundity values in H. iris, which increase markedly at shell lengths greater than 80 mm (Poore 1973). During summer, higher values for the feeding rate of the black foot paua reflect metabolic responses to increased temperature. Like Haliolis giganlea and Hcdiotis discus hannai (Ino 1943, Ino 1952). H. iris exhibited maximal grazing rates during spring and summer and the lowest values during autumn and early winter. Although this contradicts a prediction made by Poore (1969), his conclusions were based on measurements of gut fullness. Low winter temperatures may affect abalones by reducing feeding ac- tivity and increasing gut passage time. This could explain why paua with full guts were found in winter field populations. For all size groups of W. iris, consumption rates on red algae exceeded those for brown algae. Similar findings were reported by Fleming (1995a) working on Haliotis rubra. Also, differences in feeding rates between the algal species were more pronounced in summer, when the overall levels were elevated for all size groups. The assimilation efficiencies or digestibility of algae used in our experiments did not correlate with taxonomic divisions. This con- trasts with research on H. discus haimai. where assimilation effi- ciency was higher in brown than in red algae (Ino 1952, Sakai 1962). Steinberg (1985) and Sakata et al. (1988) tried to explain dis- crepancies in feeding behaviour in terms of the palatability or attractiveness of the alga. Some abalones are thought to avoid algae high in phenolic compounds because they inhibit nitrogen digestion (Fleming 1995b). Preliminary examination of phenol levels from the algae used in our experiments suggested the high- est levels for Lessonia. intermediate levels for Gracilaria. and the lowest values for the other two species (Williams 1990). If low phenol levels make an alga attractive and therefore more palatable, then subsequent choices may be made on the basis of food quality or in response to particular components such as nitrogen levels or trace elements. For example, in the case of Macrocystis. its low phenol levels, high energy content, and the ability of H . iris to assimilate it makes it potentially a better food source than some red 10- (A) Length Density = 20 3,0 2.0 Density = 80 ~m . . wm\ 1.0 ^— ^ ,0.0 i-J J-A S-0 N-D J-f 10 20 40 60 Paua Density l-J J-A S-0 N-D j-r Figure 4. Effect of paua density on individual bimonthly length in- Figure 5. Annual increments for (A) length and (B) weight of//, iris crease. held at different densities. Factors Affecting H . iris Grazing Rate 405 algae. Because it is locally abundant and readily available as drift, it could be used in abalone culture. However. Gracitaria was the best food source because it exceeded all of the other species we tested in terms of potential energy gain from algal tissue. Our grazing experiments did not provide a choice of algae; thus, we cannot comment directly about the relative attractiveness between species. However, in simple two-way choice experi- ments, Poore (1969) found preferences in the same rank order that we found for the grazing rates. These laboratory preferences, how- ever, may not operate in field populations, and examinations of gut contents have shown that H. iris will feed on the relatively unattractive L. voriegata. in the absence of other food species. This suggests that food availability, rather than food preference, is a likely determinant of the natural diet of H. iris. Several authors have suggested that abalone show consistent food preferences based on food availability in their habitat. This may partially explain some apparent inconsistencies in dietary composition, both within and between localities (Shepherd 1973. Shepherd and Steinberg 1992. McShane et al. 1994). The results of our study suggest that when the most palatable algal food is not available, paua will feed on less favourable food items. Previous knowledge is apparently not required, and paua readily accepted the branching thallus oi Gracitaria, although this alga is not nor- mally available in the Kaikoura region. The growth of H. iris in this study was low compared with estimates of abalone growth elsewhere (Day and Fleining 1992). The annual length increment in our study was approximately half that recorded from our field data and from tagged paua on the shore (Williams 1990. Pirker 1992). No explanation for this dif- ference can be provided, although the restricted monospecific diet may form a part (Wilson 1987). Recent studies on artificial diets have highlighted the value of variation in the diet, and the inclu- sion of additives such as trace elements and lipids can affect ab- alone growth (Hahn 1989. Uki and Watanabe 1992. Mai et al. 1994). Although a number of experimental field studies have investi- gated the effects of density on mollusc growth (Underwood 1979. Ortega 1985. Fletcher 1988). there have been few studies on ab- alone. McShane and Naylor (1995) recently reported similar growth rates for W. iris held in field enclosures. At densities of 0.3 and 15 m~". they concluded that there were no food or space limitations on growth. In our experiments, where the density lev- els corresponded to between 20 and 8(X) m"~. we found severely restricted growth rates of W. iris with increasing density. Although similar results were found in early studies on Haliotis giganlea discus (Ino 1943) and Haliotis cracherodii (Leighton and Bool- ootian 1963). more recent investigations into barrel culture of abalone feeding on Mucrocystis have yielded high growth rates, at densities similar to those used in our experiments (Aviles and Shepherd in press). Generally, growth rates of abalone held in cages with high water exchange and a high surface-to-volume ratio grow faster than those held in barrels or tubes. Also. Hahn ( 1989) emphasises the importance of good water quality for abalone growth, which can be inhibited by reduced oxygen levels and bacterial growth. In our experiments, although we cannot exclude water quality as a limiting factor, the detrimental effects of high density may also be due to interference (Douros 1987) or space limitations. At high density, abalone may be exposed to increased stress, resulting in higher metabolic rates (Gaty and Wilson 1986). and/or in- creased biochemical or other activity. Paua held at the two highest densities continued to feed throughout the experiment but lost weight and did not increase in overall length. However, despite these trends, a few individuals grew, as evidenced by a different shell colour. These observations suggest high individual variabil- ity in response to density stress and a remarkable ability for black foot paua to survive lengthy periods in less than favourable con- ditions. In conclusion, this study has confirmed that H iris, provided with a natural food supply, can be maintained in laboratory culture on a small scale, but we have highlighted some problems in main- taining growth rates similar to field values. Further studies are now required, using larger numbers of individuals and investigating the effects of artificial diets (Uki and Watanabe 1992) and dietary supplements (Mai et al. 1994) on feeding, energy conversion, and growth of this potentially important aquaculture species. ACKNOWLEDGMENTS We thank Allan Rodrigo and Russell Death for statistical ad- vice and Jack van Berkel. from the Edward Percival Field Station, for help in setting up the laboratory experiments and maintaining the aquarium system. We also thank Dr S. Shepherd for providing useful comments on the manuscript. LITERATURE CITED Andrew. N. L. 1986. The interaction between diet and density in Influ- encing reproductive output in the Echinoid Evccluinis chiorolicus (Val.). J. Exp. Mar. Biol. Ecol. 97:63-79, Aviles. J. G. G. & S. A. Shepherd. In press. Growth and survival of Ihe blue abalone Haliotis fulgens in barrels at Cedros Island. Baja Cali- fornia, with a review of abalone barrel culture. Aquaculture Brendelberger, H. 1995. Dietary preferences of three freshwater gastro- pods for eight natural foods of different energetic content. MalacaloRia 36:147-153. Carefoot. T, 1987, Gastropods, pp. 89-172. In: T. J. Pandian and F. J. Vemberg (Eds.l. Animal Energetics, vol 2. Academic Press, New York. Chew. K. K. 1984. Recent advances in the cultivation of molluscs in the Pacific United Stales and Canada. Aquaculture 39:69-81. Day. R. W. & A. E. Fleming. 1992. The determinants and measurements of abalone growth, pp. 141-168. In: S. A. Shepherd. M. J. Tegner & S. Guzman del Proo (eds.). Abalone of Ihe World: Biology. Fishenes and Culture. Blackwell Scientific Publications. Oxford. Douros. W. J. 1987. Stacking behaviour of an intertidal abalone: an adap- tive response or a consequence of space limitation .' J . Exp. Mar. Biol. Ecol. 108:1-14. Fleming, A. E. 1995a. Growth, intake, feed conversion efficiency and chemo!,ensory preference of Ihe Australian abalone Haliotis rubra. Aquaculture 132:297-311. Fleming, A. E. 1995b. Digestive efficiency of the Australian abalone //a/- iotis rubra in relation to growth and feed preference. Aquaculture 134:279-293. Fletcher, W. J. 1988, Intraspecific interactions between adults and juve- niles of the subtidal limpet. Patella mufria. Oecologia (Berl.) 75:272- 277. Gaty, G. & J. H. Wilson. 1986. Effect of body size, starvation, temper- ature, and oxygen tension, on the oxygen consumption of hatchery reared ormers Haliotis tuberculala. Aquaculture 56:229-227. Hahn. K. O. 1989. Nutrition and growth of abalone. pp. 135-311. In: K. O. Hahn (ed.). Handbook of Culture of Abalone and Other Marine Gastropods. CRC Press. Boca Raton, Florida. 406 Marsden and Williams Ino, T. 1943. Feeding and growth of a Japanese abalone. Haitolis giganlea discus Reeve. Bull. Jap. Soc. Sci. Fish. 1 1(5-6):17I-I74. Ino, T. 1952. Biologieal studies on tlie propagation of the Japanese aba- lone (genus Haliolis). Bull. Tokai Reg. Fish. Res. Lab. 5:29-102. Leighton, D. L. & R. A. Boolootian. 1963. Diet and growth in the black abalone Haliotis cracheriodii . Ecology 442:227-238. Mai, K., J. P. Mercer & J. Donlon. 1994. Comparative studies on the nutrition of two species of abalone, Haliotis tuberculata L. and Hali- otis discus hannai Ino. Ill Responses of abalone to various levels of dietary lipid. Aquaculture 134:65-80. McShane. P. E.. K. P. Black & M. G. Smith. 1988. Recruitment pro- cesses in Haliolis rubra (Mollusca Gastropoda) and regional hydrody- namics in southeastern Australia imply localized dispersal of larvae. J . E.xp. Mar. Bwl. Ecol. 124:175-203. McShane, P. E., H. K. Gorfine & I. A. Knockey. 1994 Factors influ- encing food selection in the abalone Haliotis rubra (Mollusca; Gas- tropoda) 7. Exp. Mar. Biol. Ecol. 176:27-37. McShane. P. E. & J. R. Naylor. 1995. Density independent growth of Haliotis iris Martyn (Mollusca: Gastropoda). J. Exp. Mar. Biol. Ecol. 190:51-60. Ortega, S. 1985. Competitive interactions among tropical intertidal lim- pets. /. Exp. Mar. Biol. Ecol. 90:11-25. Paine. R. T. & R. L. Vadas. 1969. Calorific values of manne benthic algae and their postulated relation to invertebrate food preference. Mar. Biol. (Berl.) 4:79-86. Paul, A. J., J. M. Paul, D. W. Hood & R. Neve. 1976. Observations on food preferences, daily ration requirements, and growth of Haliolis kamlschatkana Jonas in captivity. Veliger 19:303-309. Phillipson. J. 1964. A miniature bomb calonmeter for small biological samples. Oikos 15:130-139. Pickenng, T. 1986. Gracilaria-d food for cultured paua? Catch July- August (Special Paua Issue. Part II. p. 7), Pirker. J. G. 1992. Growth, shell-ring deposition, and mortality of paua [Haliotis iris Martyn) in the Kaikoura region. M.Sc. Thesis. University of Canterbury, Chnstchurch, New Zealand. 139 pp. Poore. G. C. B. 1969. The ecology of the N.Z. paua species (Molluscan). Ph.D. Thesis. University of Canterbury. Canterbury. New Zealand. 103 pp. Poore. G. C. B. 1972a. Ecology of New Zealand abalones. Haliotis spe- cies (Molluscan: Gastropoda). I -Feeding. N.Z. J. Mar. Freshwater Res. 6:11-22. Poore, G. C. B. 1972b. Ecology of New Zealand abalones. Haliotis spe- cies (Molluscan: Gastropoda) 3-Growth. A^.Z. J. Mar. Freshwater Res. 6:534-559. Poore. G. C. B. 1973. Ecology of New Zealand abalones. //900 \xm) m approximately equal numbers, with virtually no mid-size larvae. The Florida Keys have a very small spawning stock, yet densities of late-stage larvae were nearly as high as those in the Exuma Cays, where spawning stocks are large. Late-stage larvae in the Keys appear to be derived from distant spawning stocks, probably in Cuba or the western Caribbean Sea. In contrast, only 1% of the larvae in the Exuma Cays were late stages and the juvenile populations appear to depend on local spawning and recruitment. The sizes of juvenile populations were positively correlated with the mean density of late-stage larvae in both the Florida Keys (r = 0.881) and the Exuma Cays (r = 0.759), indicating the significance of larval supply in determining benthic recruitment on the local scale. The slope of the linear relationship between larval supply and juvenile population size, however, was much higher in the Exuma Cays nurseries than in Florida, suggesting important regional differences in settlement and postsettlement processes. Recruitment of benthic fauna with pelagic larvae must be considered in terms of metapopulation dynamics and both presettlemcnt and postsettlement processes. KEY WORDS: Bahamas, distribution, Florida, larval transport, length-frequency, Slromhus gigas INTRODUCTION A large proportion of benthic marine animals have "'complex life cycles" (sensu Roughgarden et al. 1988) involving pelagic eggs or larvae with high potential for dispersal as well as loss to stochastic, density-independent processes during larval develop- ment. Connell (1985) predicted that population sizes of benthic animals would be positively correlated with recruitment density where recruitment rate is low and that density-dependent mortality would rapidly destroy a direct relationship between larval settle- ment and subsequent recruitment where recruitment rate is high. Numerous studies with fishes (Sale et al. 1984, Victor 1986, Doherty 1987) and invertebrates (Keough 1984. Connell 1985, Gaines et al. 1985. Sutherland 1987. Roughgarden et al. 1988) have examined the relationship between settlement and subsequent population size. Others have made empirical analyses of the rela- tionship between larval supply and settlement and/or recruitment (Yoshioka 1982. Wethey 1984. Gaines et al. 1985. Lipcius et al. 1990. Minchinton and Scheibling 1991. Milicich et al. 1992. Peterson and Summerson 1992. Doherty and Fowler 1994). The large gastropod Strombiis gigas Linne (queen conch) forms the basis for one of the most important fisheries of the Caribbean region, with a total annual value of approximately $40 million US between 1988 and 1991 (Appeldoorn 1994). However, queen conch stocks have declined throughout the region over the past 10-20 y, and various forms of catch and size limits have been ♦Present address: National Manne Fisheries Service. Northeast Fisheries Science Center. James J. Howard Marine Sciences Laboratory. 74 Ma- gruder Road, Highlands, NJ 07732. imposed in most nations (Appeldoorn et al. 1987, Berg and Olsen 1989, Appeldoorn 1994). International trade in conch is now mon- itored by the Convention on International Trade of Endangered Species (CITES) with the hope of ensuring the species" survival. Despite complete closure of the fishery in the United States in 1985, queen conch stocks have shown little sign of recovery (Berg and Glazer 1995). This lack of recovery is poorly understood, in part because of limited knowledge of early life history, larval abundance, and recruitment processes. The ecology of the juvenile and adult queen conch is relatively well studied (Randall 1964, Weil and Laughlin 1984. Iversen et al. 1987. Stoner and Waite 1990. Stoner and Sandt 1992, Stoner et al, 1995); however, de- tailed larval descriptions for identifying the larvae of the different Strombiis species (Davis et al, 1993) and the first analyses of veliger abundance (Stoner et al. 1992, Posada and Appeldoorn 1994, Stoner et al. 1994) have appeared only recently. The re- cruitment problem is compounded by the fact that queen conch larvae spend ~3 wk in the water column and may drift hundreds of kilometers from parental stocks before settling to the benthos (Davis el al. 1993). As a result, many local populations are prob- ably replenished from distant sources, and stock management for queen conch is a multinational problem (Berg and Olsen 1989). Biochemical evidence suggests a high degree of gene flow among Caribbean populations (Mitton et al. 1989. Campton et al. 1992). This study was conducted to compare the abundance and size frequency of queen conch larvae within and between two geo- graphically distinct regions. At both sites, collections were made in nursery areas with the broadest possible range of juvenile pop- ulation size to examine the relationship between larval supply and spatial variation in recruitment. Analyses were made in the Florida 407 408 Stoner et al. Keys (United States), where populations have been heavily over- fished (Glazerand Berg 1994). and in the Exuma Cays (Bahamas), where conch populations were large and relatively stable (Stoner and Sandt 1992, Stoner and Schwarte 1994. Stoner et al. 1996). Differences in the benthic populations both within and between regions are discussed in terms of potential sources of larval re- cruitment, long-term juvenile population size, stock recovery, and management strategy. METHODS Stations in Florida Stations in the Florida Keys were chosen on the basis of long- term data for historically important conch nursery grounds in the middle and lower Florida Keys (Glazer and Berg 1994, our un- publ. data). Two stations were in nearshore locations (Tingler's Island [TI] and Big Pine Key [BP]; Fig. 1 ) that had small, ephem- eral populations of juveniles (Table 1). These sites were —1.5 m deep at mean low water (MLW), and the bottom was a mixture of macroalgae, sponges, sand, and patches of seagrass (primarily Thalassia tesludinum Konig). Two more stations were located on nursery grounds associated with shoal areas along the Florida Keys coral-reef tract, which lies ~ 10 km offshore and to the south of the islands. These areas. Delta Shoal and Looe Key National Marine Sanctuary, typically support several hundred to a few thousand conch (Table I). Delta Shoal is a shallow, coral-rubble area cov- ered with macroalgae, small corals, and patches of sand. Delta Shoal station 1 (DSl) was located in the backreef area along the northern edge of the shoal over a 1.5-m-deep platform of mixed seagrass, rock, algae and sponges. Station 2 (DS2) was -0.5 km offshore from the shoal, where depth increased rapidly from 20 to 30 m. Looe Key is composed of a very shallow coral-reef tract running east to west, with shallow rubble shoals reaching north from the ends of the reef. Station 1 ( LK I ) was behind the reef and between the shoals, where there is a shallow sand- and scagrass- covered flat. Station 2 (LK2) was -0.5 km offshore in depths of 20-30 m. At both Delta Shoal and Looe Key. juvenile conch arc found consistently on algae-covered rubble and seagrass in shal- low water. Adults are found on the reef flat at Looe Key and in the deep areas surrounding the reefs and shoals at both sites. Spawn- ing occurs principally in these deeper habitats and has not been observed north of Hawk Channel. Stations in the Bahamas Five stations were chosen on the basis of long-term data from the Lee Stocking Island area in the southern Exuma Cays (Stoner etal. 1994, Stoner etal. 1995, ourunpubl. data) (Fig. 1). Stations at Children's Bay Cay (CBC). Shark Rock (SR), Tug Boat Rock (TBR), and Neighbor Cay (NBC) were all on the shallow Great Bahama Bank to the west of the Exuma Cays (leeward in prevail- ing summer winds). The fifth station was on the windward island shelf of Lee Stocking Island in a cove off of Charlie's Beach (CHB). As is typical of large nursery grounds in the Exuma Cays, CBC and SR are characterized by moderate-density seagrass {T. tesludinum). shallow depth (3.0 m al MLW). and strong tidal currents. Both have large aggregations of 70.000 or more juvenile conch in most years (Table 1). TBR. NBC, and CHB have smaller, more ephemeral populations, with TBR being the largest and most consistent (7,000-50,000 individuals), and CHB and NBC rarely having more than 200-2,000 conch (Table 1). TBR is aroo FLORIDA BAY BIG PINE » KEY , • FLORIDA STRAIT -26- \^ -1 Florida % ^\ • 23" Cuba ^V-.. 20 ^'^^ -^^^ 82- -^.Te-'f^^^ EXUMA SOUND LEE STOCKING ISLAND CHILDREN'S BAY CAY RAT CAY Figure 1. Location of veliger-sampling sites in the Florida Keys (top) and Exuma Cays, Bahamas (bottom). Arrows in the center map .show the general locations of the two study sites. Note the scale differences for the Florida and Exuma maps. 2.0 m deep and has sparse to medium-density seagrass and strong tidal currents. Where plankton tows were made in the cove at CHB. depth is 1 .5-2.5 m and the bottom is a mixture of bare sand and patches of sparse to dense T. lestudinum. detritus, and drift algae. Conch at NBC inhabit a depth range from immediately subtidal to approximately 2.0 m in depth in an area grading from a bare sand beach to a sparsely vegetated seagrass bed (Sandt and Stoner 1993). Adult density and spawning frequency are highest in depths of 10-18 m on the island shelf to the east of the Exuma Cays (Stoner and Sandt 1992. Stoner and Schwarte 1994). Spawn- ing is relatively infrequent in bank habitats to the west of the Exuma Cays. Plankton Collections Plankton samples were collected with simple conical nets (0.5 m in diameter, 2.5 m in length, 202-jjim-pore-size mesh). Repli- Larval Supply to Conch Nurseries 409 TABLE 1. Estimates of Juvenile Population Size of Queen Conch at Nine Nurseries in the Florida Keys and Exuma Cays, Bahamas, 1988-1994. Florida Keys Exuma Cays Tingler's Delta Big Pine Looe Children's Tugboat Shark Charlie's Neighbor Island Shoal Key Key Bay Cay Rock Rock Beach Cay Year (TI) (DSD IBP) (LKll (CBC» (TBR) (SR) (CHB) (NBC) 1988 fiOO — 13.084 — — — — — 2.500 1989 1.012 — 1.716 — — — — — 6,000 1990 170 14.400 78 810 110.000 50.000 220,000 — — 1991 0 13.800 0 6,890 10.000 500 162,000 281 — 1992 536 845 3.920 722 90.000 7,000 70,000 229 — 1993 228 1 .830 368 2,116 145,900 47,600 84,900 50 100 1994 70 1.282 0 2.504 165,600 100,700 85,700 0 0 Mean 374 6,431 2,738 2.608 104,300 41,160 124,520 140 2,150 Population size at stations Tl and BP were determined by standard tag and recapture methods (Glazer and Berg 1992). whereas the larger, more dense populations In the Flonda Keys (DSl and LKI ) were surveyed by the use of standardized belt-transect methods (Glazer and Berg 1994). The numbers of juveniles in the largest populations (CBC, SR, and TBRi were estimated by measuring the density of conch in highly replicated quadrats within multiple sectors of the aggregation, which were mapped with the aid of the global positioning system (Stoner and Ray 1993. Stoner et al. 1994). cate tows ( ~ 1 .0 m • sec ~ ' ) were made at each station and sample date near the water surface. Surface sampling was appropriate because queen conch veligers are photopositive (Barile et al. 1994) and most abundant near the surface under relatively smooth sur- face conditions (Stoner and Davis 1997b). In the Florida Keys, sampling was restricted to mid-day periods when wave height was <0.5 m. In the Exuma Cays, where the nurseries are subject to strong tidal currents, sampling was restricted as described above and confined to time periods 2 h before and 1 h after the high tide. Stoner and Davis (1997a) have shown that the highest concentra- tions of veligers pass through the inlets at mid-tlood tide. We assumed, therefore, that maximum densities at the nurseries would occur close to the high tide. Tow times varied according to the concentration of plankton on the sampling date but averaged 15 min each. Tow volume was determined with a calibrated General Oceanics flow meter suspended in the mouth of the net. Nets were deployed by hand from small boats (<8 m). Plankton samples were preserved in a buffered 5% formalin-seawater mixture. At both sites, sampling was conducted between late May and late September in 1992 and 1993. seasonal periods that spanned the primary reproductive season at Lee Stocking Island (Stoner et al. 1992) and in the Florida Keys (Glazer, pers. observ.). In 1992. samples were collected every 2 wk at the Florida stations for a total of eight sets of plankton. At the Exuma stations, samples were collected every 9 d, yielding 13 sets. In 1993. the sampling effort was increased to 12 collections in the Florida Keys and 14 collec- tions near Lee Stocking Island. On each sampling date, observa- tions were made on wave height and direction, wind speed and direction, and surface-water temperature. In the laboratory, plankton samples were rinsed on a 180-p.m- pore-size mesh screen and sorted for veligers of Strombus spp. with the aid of a dissecting microscope. Even the smallest veligers of Stramlnis gigas. Slromhus costartis. and Slromhus raninus can be distinguished by use of the descriptions of Davis et al. ( 1993). These were identified to species, counted, and measured for max- imum shell length (SL). Patterns of abundance for 5. gigas were analyzed in terms of numbers of veligers per unit volume of water sampled (veligers ■ 100 m " ') for the total number of veligers and by size class. Classes used were; early-stage (<500 \x.m SL), mid-size (500-900 |jLm), and late-stage veligers (>900 |jLm). most of which were metamorphically competent. In an open system, larvae of different sizes may have different sources and size- specific density data are useful in interpreting larval production and transport processes. For example, early-stage conch veligers are only a few days old (Davis et al. 1993) and reflect local larval production, whereas late-stage veligers may have a source in more distant reproductive populations. Juvenile Population Size To test for a relationship between the supply of larvae to nurs- eries and the subsequent benthic population, regression analysis was performed using the mean seasonal density of late-stage veligers as the independent variable for each station and the esti- mated total number of juveniles in the population the following year as the dependent variable. That is, 1992 veligcr data were paired with 1993 data for juveniles, and 1993 veligcr data were paired with 1994 data for juveniles. Juveniles were predominantly 1-y-old conch, 80-120 mm SL. The methods used to estimate juvenile population size (see Table 1 ) have been described in detail for both the Florida Keys (Glazer and Berg 1992, Glazer and Berg 1994) and the Exuma Cays (Stoner and Ray 1993; Stoner et al. 1994). RESULTS Spatial Variation in Veliger and Size Frequency In 1992. only 209 queen conch veligers were collected in all 96 tows in Florida, whereas over 3.900 were collected in 130 tows in the Exuma Cays (Table 2). Total numbers collected in the two geographic areas were relatively similar in 1993 (2,300-2,700); however, most of the 2,600 veligers at LKI were newly hatched individuals collected on just one date. In Florida, the highest mean veliger concentration occurred at LKI during both 1992 (9.1 veligers • 100 m"') and 1993 (140 veligers • 100 m ') (Table 2). Over the 2-y survey period, only one queen conch veliger was collected al TI, and only two were collected at BP. Densities of conch veligers were generally higher in the Exuma Cays than in the Florida Keys. Densities near the largest populations at SR. CBC. and TBC were 15-33 veligers • 100 m ' in 1992 and 12- 410 Stoner et al. TABLE 2. Counts and Density of Queen Conch Veligers (All Stages) Collected in the Florida Keys and Exuma Cays. Bahamas, May Through September 1992 and 1993. 1992 1993 No. of Veligers Veliger Density No. of Veligers Veliger Density Site and Station Collected (no. 100 m ') Collected (no. 100 m ') Florida Keys 16 tows 24 tows Tingler's Island (Tl) 0 0 ± 0 1 0.06 ± 0.19 Delta Shoal 1 (DSD 11 0.86 ± 1.77 71 2.40 ± 6.80 Delta Shoal 2 (DS2) 29 2.60 ± 5.50 ND ND Big Pine Key (BP) 1 0.07 ± 0.27 1 0.07 ± 0.22 LooeKey 1 (LKI) 144 9.10 ± 19.8 2,637 140 ± 443 Looe Key 2 (LK2) 24 1.10 ± 1.90 ND ND Total 209 2,710 Exuma Cays 26 tows 28 tows Children's Bay Cav (CBC) 939 17,8 ± 14.7 942 12.5 ± 18.0 Tugboat Rock (TBR) 799 15.6 ± 13.7 278 3.57 ± 6.60 Shark Rock (SR) 1,576 32.5 ± 34.8 1,001 12.5 ± 9.7 Charlies Beach (CHB) 123 2.30 ± 2.70 ND ND Neighbor Cay (NBCl 494 9.80 ± 9.40 136 1.70 ± 3.30 Total 3.914 2,357 Density values are mean ± standard deviation. The number of tows made at each station is shown for each of the 2 y. ND, not determined. 36 veligers • 100 m"' in 1993 (Table 2). Densities near the smaller, more ephemeral juvenile populations were 1.7-9.7 veligers • 100 m^'. All of the queen conch veligers collected in the Florida Keys were either very small, newly hatched individuals (most were <400 |xm SL) or late-stage larvae (>900 |jim SLl. There were no intermediate stages. Nevertheless, two types of size-frequency dis- tribution were observed at Looe Key and Delta Shoal (Fig. 2). Directly over the nurseries (LKI and DSD. most of the larvae collected were early stages, probably just 1-4 d old. Conversely, no larvae < 1 .0 mm SL were collected at offshore station LK2. and only three individuals <1.0 mm were collected at DS2. The dif- ferences were highly significant for both station pairs (Kolmog- orov-Smimov tests, p < 0.01). Size-frequency distributions for veligers collected in the Ex- uma Cays were dominated by early-stage larvae (Fig. 3). At all five stations, small veligers (<500 (Ji,m SL) comprised >93% of the total numbers; however, mid- and late-stage veligers were also collected at all stations. Occurrence of Precompetent Veligers Densities of early-stage larvae were relatively high (near 25 veligers • 100 m"^) at Exuma Cays stations SR, CBC, and TBR throughout the spawning season in 1992 (Fig. 4), with strong maxima (>100 veligers • 100 m"') occurring at SR in June and late August. Although larval densities were obviously lower in 1993 than in 1992, densities of early stages were never zero at these three nurseries with large populations of juvenile conch. Among the Exuma Cays stations with low numbers of juvenile conch, early-stage veligers were constantly present at NBC in 1992, but they were present in only 4 of 13 collections made in 1993. At CHB, concentrations were low. but positive, on all but two dates with no veligers collected. None of the Exuma Cays stations showed an obvious temporal pattern in veliger density during the spawning seasons of 1992 and 1993. Among the Florida stations, early-stage larvae were suffi- ciently abundant to warrant plotting only at LKI (Fig. 4). Densi- ties were erratic, with high concentrations interspersed among zero values throughout both 1992 (50% zeros) and 1993 (33% zeros). Eighty-eight percent of all of the conch veligers collected at this station in 2 y were very recently hatched larvae (<400 (j.m SL) collected on 22 July 1993. Early-stage larvae were never collected at LK2. At DSl, early stages were collected on 4 of 20 sampling dates, with only one density estimate >1 veliger ■ 100 m" (24.8). occurring on 8 July 1993. At DS2, early-stage veligers were collected on only two dates in 1992. with values ^0.7 veligers ■ 1(K) m" \ Density of early-stage larvae showed no particular association with wind direction or speed except that none were collected on the relatively few dates (three dates at Looe Key, one date at Delta Shoal) when wind velocity exceeded 7.3 m • sec^' (15 knots). The very high concentration of newly hatched larvae estimated for LKI in July 1993 was associated with calm conditions and the highest recorded temperature for the season (32°C). Whereas no mid-stage conch veligers were collected in the Florida Keys, they were present in 56% of the samples at CBC and 44% of the samples at SR. Densities were much lower than those for early-stage larvae, typically 1 veliger • 100 m ' at TBR. NBC, and CHB. At all of the stations, mid-stage larvae were relatively uncommon before mid-June in both 1992 and 1993, probably reflecting the beginning of the reproductive season and the subsequent growth of the larvae. As with early-stage larvae, mid-stage veligers were more abundant at SR in 1992 than in 1993. At CBC, mid-stage larvae were more abundant in 1993 than in 1992. but densities were erratic in all cases. Occurrence of Competent Veligers Concentrations of late-stage larvae, all of which were compe- tent or near competent, are particularly relevant to benthic recruit- ment in the nursery habitats; therefore, these were considered sep- Larval Supply to Conch Nurseries 411 70 60 50 40 30 20 10 0 T I ■^^^^^^^^^^^^^~~ LK1 (n = 2774) 300 400 500 600 —I — I — I — I — T — r — I — I — I — I — 700 800 900 1000 1100 — I 1 r 1 1 1 1200 1300 a z o c o 3 o 70 60 50 40 30 20 10 ■ LK2 (n = 24) -! — r — T — I — I — r — 1 — 1 — I — 1 — I — I — r — i — r — r — i — i — i — i — r~T — i — i — r — i — i — i — i 300 400 500 600 700 aOO 900 1000 1100 1200 1300 7D r 60 ■ 50 40 30 - y 20 - 1 10 - I n n 70 300 60 - 50 - 40 - 30 - 20 - 10 - 0 - , , ) DS1 (n = 82) n — ^ — r — I — I — I — I — 1 — I — : — I — T — 1 — I — I — I — r~ 300 400 500 600 700 800 900 1000 DS2 (n = 29) 1100 1200 1300 1100 1200 1300 Shell Length (um) Figure 2. Length-frequency distribution of queen conch veligers at four stations in the Florida Keys in 1992 and 1993. The total number of larvae shown for station LKI is slightly less than that reported in Table 2 (2,781) because some shells were damaged and not measured. arately (Table 3; Fig. 6). The highest concentrations of late-stage larvae occurred in 1992. at the offshore Florida stations DS2 ( 1 .34 veligers • 100 m"^) and LK2 (0.85 veligers • 100 m"'). where late stages comprised most of the larvae (Fig. 2). Among the nursery sites, the densities of late-stage larvae ranged from zero at stations characterized by very small populations of juveniles. Tl and BP. to 0.34 veligers • 100 m"' at DSl in 1993 (Table 3). Mean concentrations were relatively consistent between 1992 and 1993 at the Florida nurseries; however, the frequency of occur- rence of late-stage veligers was lower in the second year. The highest concentrations of late-stage veligers at nursery sites LK 1 and DSl were associated with onshore winds from the south. Conversely, at all four stations near the reef tract (at Delta Shoal and Looe Key), larval densities were zero whenever the wind was north of east (<90-110° true), suggesting that the larvae were transported on and off the shelf with the surface layer. In general, the mean densities of late-stage larvae at the Exuma Cays stations were not much higher than those in the Florida Keys (Table 3). despite very large differences in total larval densities (Table 2). The highest density for one date was 5.9 veligers • 100 m ' at CBC in August 1993; however, densities >2 veli- gers ■ 100 m~' were uncommon (Fig. 6). Late-stage larvae oc- curred sporadically, with less than half of the sampling dates yield- ing late-stage larvae (Table 3). and there was no concordance in the abundance patterns at SR and CBC, which lie in adjacent tidal flow fields and are separated by just 7 km. Consistent with the pattern observed for early- and mid-stage larvae, late-stage larvae were more abundant at SR in 1992 than in 1993. 4i: Stonfr et al. 70 60 50 40 30 20 10 0 SR (n = 2576) 20 1 5 1 0 05 \- 00 300 400 — I — I— 500 600 700 800 900 1000 1100 1200 1300 o >. o c 0) 3 a> 70 60 50 40 30 20 10 0 70 60 50 40 30 20 10 0 CBC (n = 1887) 300 400 TBR (n = 1077) 500 600 20 15 10 05 0.0 300 400 — I — r— 500 —I — I — r— — I — I — r- 700 800 600 — 1 1 — 700 900 — (Ill 1 1 r 1 1 1 1 1 1 1 1000 1100 1200 1300 500 600 700 800 900 1000 1100 1200 1300 — I 1— T T 1 1 1 1 t~~i r 1 1 1 1 1 1 1 1 1 r 1 800 900 1000 1100 1200 1300 700 500 600 700 800 900 1000 1100 1200 1300 800 900 — r~T 1 — I r — 1 — I — r — I — T — I — I — 1 — I 1000 1100 1200 1300 1 T T — r — 1 — I — I — 1 — r — r — i — i — r — i — i — i — i — i — r — r — i — t — r — i — r — ^ — r — i — i — r — i — r — i i 500 600 700 800 900 1000 1100 1200 1300 700 800 900 — I 1 1 1 1~^ r 1 1 1 1 1 1 1 1000 1100 1200 1300 Shell Length (um) Figure 3. Length-frequency distribution of queen concli veligers at five stations in the Exuma Cays in 1992 and 1993. The frequency distri- butions of larvae >500 |xm SL are show n in the Insets. Note the scale difference on the y-axes. Small differences in the total numbers reported here and in Table 1 occur because some shells were damaged and not measured. Larval Suppi v to Conc h Nurseries 413 125 100 75 50 25 0 125 100 75 50 h 25 0 125 100 75 50 25 0 125 100 75 50 25 0 125 100 75 50 - 25 - 0 ./vA-<%^ • • •i •-■ •--••m »» ■ •• • m^^~^ -• — •- ^ o s 8 0) a 00 s- c 3 C 3 3 3 < O) 3 s- ^ -5 -J < w 1992 125 100 75 50 25 - 125 100 75 50 25 0 125 100 75 50 25 0 125 100 75 50 25 0 125 100 75 50 25 0 125 100 75 50 25 h SR M,^* ^.%Vv% TBR NBC u*:^^ CHB No data ^> 1 1610 ±52 LK1 -<••' • • CM o o p a> o) 00 00 2 -5 -> ^ < w 1993 Figure 4. Density of newly hatched queen conch veligers (500 |jini SI,( at five nurseries in the Exuma Cays, Bahamas, and at the Looe Key nursery (Florida Keys) in 1992 and 1993. Numbers of larvae collected at the other study sites were too low to plot. Values shown are mean ± standard error (n = 2). 414 E 8 o o 6 ^v ■**• o 4 c 0) (Q 2 > k. 5 0 0) O) 10 CO *« (/> H ■o s 6 o >> ♦^ 4 (0 c 0> Q 2 Signer et al 10 8 • rnrn^^i^*-^' >- •- o » "^ ^ 2 ff ^ Q O O) <7> 00 CD c "5 3 O) Q. i^ 3 ^ < 3 a O -J ^ < w 2 - SR • •• ■*•' * u B ■ 6 ■ 4 ■ 2 CBC ^ c^ •- rt S 5 <5 3 3 ^ O) s 00 3 O) Q. -t 3 <]> < W o O 1992 1993 Figure 5. Density of mid-stage queen conch veligers (500-900 jim SL) at two nurseries in the Exuma Cays, Bahamas, in 1992 and 1993. Numbers of larvae collected at the other study sites were too low to plot. Values shown are mean ± standard error (n = 2). TABLE 3. Counts and Density of Late-Stage Queen Conch Veligers Collected in the Florida Keys and Exuma Cays, Bahamas, May Through September 1992 and 1993. 1992 1993 No. of % of No. of % of Veligers Collections Veliger Density Veligers Collections Veliger Density Site and Station Collected With Veligers (no. 100 m ') Collected With Veligers (no. 100 m '» Florida Keys 16 tows 24 tows Tingler's Island (Tl) 0 0 0 ± 0 0 0 0 ± 0 Delta Shoal 1 {DSD 6 50 0.28 ± 0.34 5 8 0.34 ± 1.04 Delta Shoal 2 (DS2) 26 50 1.34 ± 1.82 ND ND ND Big Pine Key (BP) 0 0 0 ± 0 0 0 0 ± 0 Looe Key 1 (LKl) 2 25 0.10 ± 0.21 5 17 0.16 ± 0.56 LooeKev 2 (LK2) 24 38 0.85 ± 1.65 ND ND ND Total 58 10 Exuma Cays 26 tows 28 tows Children's Bay Cay (CBC) 8 15 0.32 ± 0.79 54 50 0.55 ± 1.55 Tugboat Rock (TBR) 6 15 0.13 ± 0.38 14 36 0.21 ± 0.34 Shark Rock (SR) 23 39 0.52 ± 1.45 6 29 0.08 ± 0.14 Charlie's Beach (CHB) 2 8 0.06 ± 0.23 ND ND ND Neighbor Cay (NBC) T 15 0.04 ± 0.09 10 43 0.11 ± 0.17 Total 41 84 Density values are mean ± standard deviation. The number of tows made at each station is shown for each of the 2 v ND, not determined Larval Supply to Conch Nurseries 8 415 S 05 3 a> Q. < 3 0) < W 1993 0) g] oo □) *" 3 O) a. < 3 o < W 1992 Figure 6. Density of late-stage queen conch vcligers (>900 jjim SL) at two stations in the Exuma Cays and at two stations in the Florida Keys in 1992 and 1993. Relatively few late-stage larvae were collected at the other sampling stations (see Table i). Values shown are mean ± standard error (n = 2). 416 Stoner et al. Relationships Between Veligers and Juvenile Population Size Two general observations can be made about the relationship between the density of veligers and juvenile populations at the study sites. First, veliger densities were consistently low (Table 2) at stations with ephemeral or small juvenile populations, such as Tl and BP in the Florida Keys and CHB in the Exuma Cays (Table 1). Second, in the Exuma Cays, veligers were always present, and density maxima were high at nurseries where there were consis- tently large aggregations of juvenile conch (i.e., CBC and SR). Similarly, the highest larval densities in the Florida Keys occurred at LKl and DSl, where juveniles were most abundant. There was a close, positive correlation between larval supply (mean seasonal density of late-stage conch veligers) and the size of the juvenile population 1 y later, both in the Exuma Cays (r = 0.759; p = 0.018) and in the Florida Keys (r = 0.881, p = 0.004) (Fig. 7). Analysis of covariance, however, indicated that the slopes of the regression lines for the two regions were different (F,| ,,, = 5.061; p = 0.042). It is clear from the plots (Fig. 7) that the slope for the Exuma Cays stations was much higher than that for the Florida Keys; the difference was >40 times. For ex- ample, a density of 0.3 late-stage veliger ■ 100 m ' in the Florida Keys was associated with —2,000 juvenile conch, whereas the same concentration of larvae in the Exuma Cays was associated with -80.000 juveniles. There was also a significant positive correlation between the abundance of juvenile conch in Florida nursery grounds and the percentage of plankton sampling dates (during the previous year) that yielded late-stage veligers (r = 0.733; p = 0.039). The correlation was not significant in the Exuma Cays (r = 0.410; p = 0.273). DISCUSSION Veliger Size Frequency and Probable Sources Detailed length-frequency data for queen conch veligers gen- erated in this study provide important insights into the different 0 01 0 2 03 04 05 06 0 01 02 03 04 Of Competent Veligers (no. / 100 m') Figure 7. Relationship between the mean density of late-stage queen conch larvae at a nursery ground and the size of the benthic juvenile population in the subsequent year. The relationships are shown for nurseries in the Exuma Cays, Bahamas, and Florida Keys. Circles represent the larval collections for 1992 and juvenile surveys in 1993. Triangles represent larvae in 1993 and juveniles in 1994. Linear re- gressions and 95% confidence intervals are shown. mechanisms of recruitment and sources of larvae for the two study areas. At all stations except the offshore non-nursery sites DS2 and LK2 in the Florida Keys, the majority of veligers were <500 jxm SL. On the basis of growth curves provided by Davis et al. (1993), these small veligers were no more than 5-6 d old and must have had a local source. With the exception of one collection made in Looe Key National Marine Sanctuary in 1993. the density of newly hatched conch larvae was very low compared with the den- sities of similar sized veligers in the Exuma Cays, Bahamas. The difference relates to the abundance of spawners in the two areas. Adult conch in the Keys were seriously depleted by overfishing and have not recovered since fishing was ended in 1985 (Glazer unpubl. obs.). In 1992, there were only -6,000 adult conch in the 200-km-long island chain of the Florida Keys from Carysfort Reef to Western Dr\' Rocks. Many of these adults were found in Looe Key National Marine Sanctuary, where the highest concentrations of early-stage veligers were collected in both 1992 and 1993. In contrast with the low numbers of reproductive conch in the Florida Keys, densities of adults near Lee Stocking Island ranged from 2 to 88 individuals/ha in 1991. with an estimate of 89.000 adults in just a 12-km-long section of the island shelf (Stoner and Schwarte 1994). This provides a plausible explanation for the large numbers of early-stage larvae collected in the Exuma Cays compared with the low densities in Florida. An analogous rela- tionship between the abundance of early-stage larvae and adult densities has been described for fished and unfished areas in the Exuma Cays island chain (Stoner and Ray. unpub. obs). Temporal variation in the densities of early-stage larvae is in- fluenced by local spawning frequency, egg hatching, and physical factors such as sea surface conditions. The high abundance of newly hatched veligers at LKl on 22 July 1993, for example, was associated with very warm water temperature (32°C), known to influence spawning ( Stoner etal. 1992). Stoner and Davis ( 1997b) have shown that conch larval abundance in the upper water column is influenced by wave action, and calm conditions probably al- lowed conch larvae to accumulate both in the surface layer and in backreef areas such as that near Looe Key reef. However, it is impossible thus far to separate the effects of spawning, hatching, and larval behavior and transport on larval abundance patterns. Densities of mid- and late-stage larvae are more relevant than those of early stages to the recruitment process. The complete lack of mid-size veligers in the Florida Keys and the high abundance of late-stage larvae relative to early stages, particularly in the off- shore sites, indicate that the source for these late stages was prob- ably not local. It is possible that the late-stage veligers were spawned in Florida and retained in gyres south of the Keys (Lee et al. 1992. Lee et al. in press). This retention mechanism has been hypothesized for lobsters in the genus Scyllanis. which have a 1- to 2-mo larval phase (Yeung and McGowan 1991); however, two lines of evidence indicate that the retention of conch larvae in the Florida Strait is unlikely. First, no intermediate-size larvae have ever been collected in the waters of the Florida Keys or Florida Strait (see below), and second, the densities of late-stage veligers were equal to or higher than those for early-stage larvae. Rates of mortality for queen conch larvae are unknown but assumed to be high. Given distances, and average current patterns and velocities between the Yucatan Strait and the Florida Strait, coupled with the assumed age of veligers collected in the Florida Keys, it is most likely that late-stage veligers were transported from spawning pop- ulations in Cuba, Mexico, or Belize. Such a larval transport mech- Larval Supply to Conch Nurseries 417 anisni has been assumed tor spiny lobster [Puniilnus spp. ) ( Yeung and McGowan 1991) and postulated for queen coneh (Berg and Olsen 1989. Mitton et al. 1989. Campton et al. 1991. Davis et al. 1993). Long-distance transport is well documented for a variety of marine molluscs (Scheltema 1971. Scheltema 1986). Circumstantial evidence supporting the hypothesis that surface currents carry queen conch larvae from the Caribbean Sea to the Straits of Florida was provided in recent collections made m the Florida Current. In June 1993, a mean density of 8.1 veligers • 100 m ~ "* was found 35 km south of Delta Shoal (Stoncr unpubl. obs.). All of the veligers collected were >1 .0 mm SL and near nietamorphic competence. This concentration is an order of magnitude higher than most values found within the Keys, and only one collection, made late in the spawning season at DS2. yielded a higher concentration of late-stage larvae. Given assumed (high) natural mortality rates in conch veligers. it is improbable that high concentrations of late-stage larvae originated in the adult- poor Keys environment, where newly hatched larvae are relatively uncommon. Support for the hypothesis that larvae drift from Cuba to the Florida Keys has been provided recently by the release of surface drifters along the north shore of Cuba (T. Lee pers. com- mun.). The drogues made direct paths from Cuba to Florida over several days. Under prevailing conditions, the Florida Current front is 10-20 km south of the reef tract, and exchange between the Keys and the current may be uncommon, as shown by the sporadic presence of late-stage larvae at Delta Shoal and Looe Key. The Exuma Cays island chain is probably a more efficient system than the Florida Keys in maintaining high concentrations of conch larvae close to the island shelf and nursery grounds. During the summer, when conch spawn, winds are nearly always onshore (east to southeast), and the prevailing northwest current (6-12 cm ■ sec"') along the Exuma Cays has a significant onshore (cross-shelf) component (N. P. Smith pers. commun.). The larvae are then drawn onto the nursery grounds of the Great Bahama Bank through the island passes by strong tidal currents (Stoncr and Davis 1997a). A net flow of water onto the Bank in the pass north of Lee Stocking Island has been observed (Smith and Stoner 1993). and larval concentrations are often higher in nursery areas than offshore near the spawning grounds (Stoner et al. 1992. Stoner and Davis 1997a). The Relationship Between Larval Supply and Juvenile Populations Although cause and effect are not established in a descriptive study, differences in larval supply measured in this investigation provide a plausible explanation for the observed differences in juvenile populations of queen conch in both the Florida Keys and the Exuma Cays. This is consistent with Connell's ( 1985) sugges- tion that population size will be correlated with recruitment at low recruitment densities. Similar patterns of spatial variation related to larval supply have been observed recently for coral reef fishes (Milicich et al. 1992, Doherty and Fowler 1994) and barnacles (Bertness et al. 1992). Regardless of the exact source of late-stage larvae for Florida Keys nursery areas, the correlation between larval abundance and juvenile population size in Florida was very high, with just one point lying outside a nearly perfect linear expression. It now ap- pears that ephemeral and small populations of queen conch that exist close inshore along the islands are limited by a general lack of larvae reaching these nursery grounds. Only one veliger (early stage) was collected north of Hawk Channel, where a westward tlowing current (N. P. Smith, pers. commun.) may effectively bar the transport of larvae from local spawning grounds, all found along the reef tract. The correlation between larval abundance and subsequent ju- venile population size was also significant in the Exuma Cays, but the relationship was different from that observed in the Florida Keys in two ways. The correlation coefficient was lower in the Exumas, and the slope of the regression was much higher, illus- trating that fact that deviation from a linear model of the relation- ship can vary both within and between sites. Some of the deviation from the linear relationship among nurs- ery stations in the Exuma Cays can be explained by differences in larval delivery rates. The density of larvae does not measure the actual availability of larvae to a site, and flow past the settlement substratum must be considered (Olmi et al. 1990. Yund et al. 1991 ). In relative terms, measurements of larval density underes- timate larval supply to sites with high flows and overestimate supply at sites with low flows. For example, current velocities at the nearshore stations (CHB and NBC) were relatively low and both had juvenile populations falling below the regression line. Nurseries with the highest tidal current velocities (SR and CBC) had juvenile abundances above the regression line. Attempts to collect queen conch veligers in tube traps, which integrate the abundance of larvae reaching a site over time (Yund et al. 1991 ). have not been successful, even at station SR. where veligers were most abundant (Stoner unpubl. obs.). Undoubtedly, recruitment to the benthos involves a complex interaction of the density of po- tential settlers and the regularity and rate of their arrival. The most remarkable difference in the relationship between mean density of late-stage larvae at a nursery ground and the subsequent juvenile population size in Florida and the Exuma Cays was the difference in slopes. Relatively similar mean densities of late-stage larvae were associated with very much higher juvenile populations in the Exuma Cays; the difference was as high as 40 limes. Several explanations for this difference are plausible: (1) Rale of delivery — The most productive nurseries in the Exuma Cays are characterized by high current velocities (Stoner et al. 1994. Stoner et al. 1995). with much lower tidal velocities at the Florida Keys nurseries. (2) Frequency of larval delivery — Generally, the Exuma sites had higher frequencies of larval deliv- ery to the nurseries than those in Florida, particularly in 1993. (3) Si:e (if the suitable nursery habitats — In the Exuma Cays, habitats suitable to queen conch are large and typically have been below their carrying capacity for juvenile conch (Stoner et al. 1994, Ray and Stoner 1994). This provides a large potential settlement area for arriving larvae and may increase the actual numbers of larvae settling. Suitable nurseries in the Florida Keys are associated with relatively specific small backreef and nearshore habitats, more analogous to the small nearshore nurseries in the Exuma Cays (e.g., NBC and CHB) than to the large, open seagrass nurseries of the Great Bahama Bank (e.g., TBR, SR, and CBC). (4) Settlement cues — Recent experiments designed to test the response of late- stage queen conch larvae to natural cues from known nursery habitats near Lee Stocking Island (Davis and Stoner 1994) and in the Florida Keys (Stoner et al. unpub. obs.) have shown that settlement and nietamorphic responses to sediments and macro- phytes in the Exuma Cays are stronger than the responses to anal- ogous substrata in the Florida Keys. Therefore, the frequency of settlement in the Florida Keys may be low. (5) Postsettlement processes — Growth and survivorship in the period after settlement 418 Stoner et al. can have a very large influence on the numbers of benthic juve- niles animals in a benthic population (Keough and Downes 1982. Luckenbach 1984, Rowley 1989, Keesing and Halford 1992). Mortality rates in young queen conch are highly site specific and inversely density dependent (Ray and Stoner 1994). No compara- ble experiments have been conducted to compare mortalities be- tween the two sites; however, the small size and low density of juvenile conch densities in Florida may prevent the safety in num- bers observed in conch nurseries in the Exuma Cays. It is very likely that all of these mechanisms play at least some role in the low recruitment success of queen conch populations in the Florida Keys after 10 y of fishing moratorium. Russ (1991) postulated that natural variation in spawning out- put directly affects local population size in most marine species because of high fecundities and broad dispersal capabilities. How- ever, mtense fishing pressure on fishes and invertebrates in the Caribbean has reduced the mean size and abundance of these spe- cies in many regions, and some self-recruiting systems with in- tensive local fisheries may be vulnerable to limitations related to reproductive output and larval supply (Munro et al. 1973, Munro 1983). Overfishing may, in fact, explain the apparent lack of recovery in Florida Keys populations. In the Florida Keys, adult conch populations were probably harvested to the point of recruit- ment overfishing by the mid-19S0s. when a fishing moratorium was established. Today, the populations appear to depend on spo- radic influxes of larvae from the Florida Current. If this form of recruitment is not effective in delivering a regular supply of larvae to the Florida Keys, the rehabilitation of queen conch stocks may depend on the success of hatcheiy rearing and the release of cul- tured juveniles. However, the limitations of releasing hatchery stocks are well known (Appcldoom and Ballantine 1983. Jory and Iversen 1983. Stoner 1994, Stoner and Davis 1994). Future studies will need to determine the relative significance of different larval sources in the Florida Keys and under what conditions larvae in the Florida Current can recruit to the Keys. Most important, species with pelagic larvae and high dispersal potential will need to be managed from the standpoint of meta- population dynamics rather than on the basis of local populations (Farmer and Berg 1989. Fairweather 1991, Shepherd and Brown 1993. Man et al. 1995). Unfortunately, clear genetic markers have not been found for queen conch (Mitton et al. 1989). and other biochemical markers or new techniques are needed to identify stocks and stock sources. Models integrating oceanographic pro- cesses with larval production and behavior may provide another means of answering the important question of larval source. In any case, understanding larval recruitment processes and international cooperation related to spawning-stock maintenance will be crucial to the wise management of queen conch and other fishery re- sources in the greater Caribbean region. ' ACKNOWLEDGMENTS This research was supported by a grant from the National Un- dersea Research Program of NOAA (U.S. Department of Com- merce) and the Shearwater Foundation (New York) to the Carib- bean Marine Research Center and by funding provided for conch research by the Florida Department of Environmental Protection. The Looe Key National Marine Sanctuary provided boat time for sampling near Looe Key. L. Anderson, D. Barile. B. Bower- Dennis. A. Dalton. C. Harnden, J. Lally, S. O'Connell, M. Ray, and J Walsh assisted in sample collecting and sorting. M. Davis provided species confirmations and measurements. D. Forcucci of the Florida Institute of Oceanography provided the meteorological data for Sombrero Key. and H. Proft assisted with analysis of meteorological data. This study profited from discussion with N. Smith and T. Lee about the physical oceanography of the two study areas. M. Davis and M. Ray provided helpful criticism of the manuscript. LITERATURE CITED Appeldoom. R. S. 1994. Queen conch management and research: status. needs, and priorities, pp. 301-319. In: R. S. Appeldoom and B. Ro- driguez (eds.). Queen Conch Biology, Fisheries, and Mariculture. Fundacion Cientifica Los Roques, Caracas. Venezuela. Appeldoom. R. S. & D. L. Ballantine. 1983. Field release of cultured queen conchs in Puerto Rico: implications for stock restoration. Proc. GulfCaribb. Fish. hut. 3.^:89-98. Appeldoom. R. S., G. D. Dennis & O. Monterrosa-Lopez. 1987. Review of shared demersal resources of Puerto Rico and the lesser Antilles region. FAO Fish. Rep. 383:36-106. Barile. P. J., A. W. Stoner& C. M. Young, 1994, Phototaxis and vertical migration of the queen conch {Siromhus gigus Linnel veliger larvae. J. Exp. Mar. Biol. Ecol. 183:147-162. Berg, C. J., Jr. & R, A, Glazer. (1995). Stock assessment of a large marine gastropod (Sirombus gigas) using randomized and stratified towed-diver censusing, ICES Marine Science Symposium. 199:247- 258. Berg, C. J., Jr. & D. A. Olsen. 1989. Conservation and management of queen conch (Sirombus gigas) fisheries in the Caribbean, pp. 421-442. In: J. F. Caddy (ed.). Marine Invertebrate Fisheries: Their Assessment and Management. Wiley. New York. Bertness, M. D., S. D. Gaines, E. G. Stephens & P, O, Yund, 1992, Components of recruitment in populations of the acorn barnacle Semibalanus halanoides (Linnaeus). J. E.xp. Mar. Biol. Ecol. 156: 199-215. Campton, D. E., C, ,1, Berg. Jr.. L, M, Rohi.son & R, A, Glazer, 1992, Genetic palchiness among populations of queen conch, Sirombus gi- gas. in the Florida Keys and Bimini. Fish. Bull. U.S. 90:250-259, Connell, J, H. 1985. The consequences of variation in initial settlement vs, postsettlement mortality in rocky intertidal communities. J. E.xp. Mar. Biol. Ecol. 95:11-15, Davis. M,. C. A, Bolton & A, W. Stoner, 1993. A comparison of larval development, growth, and shell morphology in three Caribbean Srrom- bus species. Veliger 36:236-244. Davis, M. & A. W. Stoner. 1994. Trophic cues induce metamorphosis of queen conch larvae (Sirombus gigas Linnaeus), J. E.xp. Mar. Biol. Ecol. 180:83-102, Doherty, P. 1987, The replenishment of populations of coral reef fishes, recruitment surveys, and the problems of variability manifest on mul- tiple scales. Bull. Mar. Sci. 41:411-422, Doherty, P, & T, Fowler, 1994, An empirical test of recruitment limitalion in a coral reef fish. Science 263:935-939. Fairweather, P. G, 1991, Implications of "supply-side" ecology for en- vironmental assessment and management. Trends Ecol. Evol. 6:60-63. Farmer, M. W. & C. J. Berg, Jr. 1989. Circulation around islands, gene flow, and fisheries management. Proc. Gulf Caribb. Fish. Insi. 39: 318-330. Gaines. S,, S. Brown & J, Roughgarden, 1985, Spatial variation in larval concentrations as a cause of spatial variation in settlement for the barnacle, Balanus glandula. Oecologia 67:267-272, Glazer. R. A. & C, J, Berg, Jr, 1992, Growth and mortality of the queen conch, Sirombus gigas, in Florida, Proc. Gulf Caribb. Fish. Inst. 42:153-157, Larval Supply to Cowrn Nurseries 419 Glazer, R. A. & C. J. Berg. Jr. 1994 Queen conch research in Florida: an overview, pp. 79-93. In: R. S. Appeldoom and B. Rodriquez (eds.). Queen Conch Biology, Fishenes and Mariculture. Fundacion Cienti- fica Los Roques, Caracas. Venezuela. Iversen. E. S.. E. S. Rutherford. S. P. Bannerol & D. E. Jory. 1987. Biological data on Berry Islands (Bahamas) queen conchs, Strombus gigas. with mariculture and fisheries management implications. Fish. Bull. U.S. 85:299-310. Jory. D. E. & E. S. Iversen. 1983. Queen conch predators: not a road- block to mariculture. Proc. Gulf. Caribb. Fish. Insl. 35:108-1 1 1 Keough, M. J. 1984. Dynamics of the cpifauna of the bivalve Pinna bicolor: interactions among recruitment, predation, and competition. Ecology 65:846-857. Keough, M. J. & B. J. Downes. 1982. Recruitment of marine inverte- brates: the role of active larval choices and early mortality. Oecologica 54:348-352. Keesing, J. K. & A. R. Halford. 1992. Importance of postsettlement pro- cesses for the population dynamics of Acanrhasier plani (L.). Aiisf. J. Mar. Freshwater Res. 43:635-651. Lee, T. N., M. E. Clarke, E. Williams, A. F. Szmant, and T. Berger. 1994. Evolution of the Tortugas Gyre and its influence on recruitment in the Flonda Keys. Bull. Mar. Sci. 54:621-646. Lee, T. N., C. Rooth, E. Williams, M. McGowan, A. F. Szmant & M. E. Clarke. 1992. Influence of Florida Current, gyres and wind-dnven circulation on transport of larvae and recruitment in the Florida Keys coral reefs. J. Conlin. Shelf Res. 12:971-1002. Lipcius. R. N.. E. J. Olmi & J. Van Montfrans. 1990. Planktonic avail- ability, molt stage and settlement of blue crab postlarvae. Bull. Mar. Sci. 46:230-243. Luckenbach, M. W. 1984. Settlement and early post-settlement survival in the recruitment of Mulinia lateralis (Bivalvia). Mar. Ecol. Prog. Ser. 17:245-250. Man, A., R. Law & N. V. C. Polunin. 1995. Role of marine reserves in recruitment to reef fisheries: a metapopulation model. Biol Con.serw 71:197-204. Minchinton, T. E. & R. E. Scheibling. 1991. The influence of larval supply and settlement on the population structure of barnacles. Ecology 72:1867-1879. Milicich, M. J., M. G. Meekan & P. J. Doherty. 1992. Larval supply: a good predictor of recruitment of three species of reef flsh (Pomacen- tridae). Mar. Ecol. Prog. Ser. 86:153-166. Mitton, J B , C J. Berg, Jr & K. S. Orr. 1989 Population structure, larval dispersal, and gene flow in the queen conch. Strombus gigas. of the Caribbean. Biol. Bull. 177:356-362. Munro, J. L. (ed.). 1983. Caribbean coral reef fishery resources. ICLARM Stud. Rev. 7:1-276. Munro, J. L., V. C. Gaut, R. Thompson & P. H. Reeson. 1973. The spawning seasons of Caribbean reef fishes. J Fish Biol. 5:69-84. Olmi, E. J., Ill, J. Van Montfrans, R. N. Lipcius, R. J Orth & P. W. Sadler. 1990. Variation in planktonic availability and settlement of blue crab megalopae in the York River, Virginia. Bull. Mar. Sci. 46:230-243. Peterson, C. H. & H. C. Summerson. 1992. Basin-scale coherence of population dynamics of an exploited marine invertebrate, the bay scal- lop: implications of recruitment limitation. Mar. Ecol. Prog. Ser. 90: 257-272. Posada, J. & R. S. Appeldoom. 1994. Preliminary observations on the distribution of Strombus larvae in the eastern Canbbean. pp. 191-199. In: R. S. Appeldoom and B. Rodriquez (eds. ). Queen Conch Biology, Fisheries and Mariculture. Fundacion Cientifica Los Roques, Caracas, Venezuela. Randall, J. E. 1964. Contributions to the biology of the queen conch, Strombus gigas. Bull. Mar. Sci. Gulf Caribb. 14:246-295. Ray. M. & A. W. Stoner. 1994. Experimental analysis of growth and survivorship in a marine gastropod aggregation: balancing growth with safety in numbers. Mar. Ecol. Prog. Ser. 105:47-59. Roughgarden. J., S. D. Gaines & H. Possingham. 1988. Recmilment dynamics in complex life cycles. Science 241:1460-1466. Rowley, R.J. 1989. Settlement and recmitment of sea urchins iStrongy- loccntrotus spp.) in a sea-urchin barren ground and a kelp bed: are populations regulated by settlement or post-settlement processes'.' Mar. Bu)l. 100:485^94. Russ, G. R. 1991. Coral reef fisheries: effects and yields, pp. 601-636. In: P. F. Sale (ed.). The Ecology of Fishes on Coral Reefs Academic Press, New York. Sale, P. F., P. J. Doherty, G. I. Eckert, W A. Douglas & D. J. Ferrell. 1984. Large scale spatial and temporal variation in recmitment to fish populations on coral reefs. Oecologia 64:191-198. Sandt, V. J. & A. W. Stoner. 1993. Ontogenetic shift in habitat by eariy juvenile queen conch, Strombus gigas: pattems and potential mecha- nisms. Fish. Bull. U.S. 91:516-525. Schellema, R. S. 1971. Larval dispersal as a means of genetic exchange between geographically separated populations of shoal-water benthic manne gastropods. Biol. Bull. 140:284-322. Scheltema, R. S. 1986. Long distance dispersal by planktonic larvae of shoal-water benthic invertebrates among central Pacific islands. Bull. Mar. Sci. 39:241-256. Shepherd, S. A. & L. D. Brown. 1993. What is an abalone stock: impli- cations for the role of refugia in conservation. Can. J. Fish. Aquat. Sci. 50:2001-2009. Smith, N. P. & A. W. Stoner. 1993. Computer simulation of larval trans- port through tidal channels: role of vertical migration. Estuar. Coastal Shelf Sci. 37:43-58. Stoner, A. W. 1994. Significance of habitat and stock pre-testing for en- hancement of natural fisheries: experimental analyses with queen conch. J. World Aquacult. Soc. 25:155-165. Stoner, A. W. & M. Davis. 1994. Experimental outplanting of juvenile queen conch {Strombus gigas): comparison of wild and hatchery-reared stocks. Fish. Bull. U.S. 92:390^11. Stoner, A. W. & M. Davis. 1997a. Abundance and distnbution of queen conch veligers {Strombus gigas) in the central Bahamas: 1. Horizontal pattems in relation to reproductive and nursery grounds. Bull. Mar. Sci. (in press). Stoner, A. W. & M. Davis. 1997b. Abundance and distribution of queen conch veligers {Strombus gigas Linne) in the central Bahamas: 2. Ver- tical pattems in nearshore and deep-water habitats. Bull. Mar. Sci. (in press). Stoner, A. W., M. D. Hanisak, N. P. Smith & R. A. Armstrong. 1994. Large-scale distribution of queen conch nursery habitats: implications for stock enhancement, pp. 169-189. In: R. S. Appeldoom and B. Rodriguez (eds.). Queen Conch Biology, Fisheries, and Mariculture. Fundacion Cientifica Los Roques, Caracas, Venezuela. Stoner, A. W., J. Lin & M. D. Hanisak. 1995. Relationships between seagrass bed characteristics and juvenile queen conch {Strombus gigas) abundance in the Bahamas. J. Shellfish Res. 14:315-323. Stoner. A. W.. P. L. Pitts & R. A. Armstrong. 1996. The interaction of physical and biological factors in the large-scale distribution of juvenile queen conch in seagrass meadows. Bull Mar. Sci. 58:217-233. Stoner. A. W. & M. Ray. 1993, .Aggregation dynamics in juvenile queen conch {Strombus gigas): population stmclure, mortality, growth, and migration. Mar. Biol. 116:571-582. Stoner, A. W. & V. J. Sandt. 1992. Population stmcture, seasonal move- ments and feeding of queen conch, Strombus gigas, in deep water habitats of the Bahamas. Bull. Mar. Sci. 51:287-300. Stoner, A. W.. V. J. Sandt & I. F. Boidron-Metairon. 1992. Reproduc- tive activity of queen conch. Strombus gigas. and the abundance of veligers. Fish. Bull. U.S. 90:161-170. Stoner, A. W. & K. C. Schwarte. 1994. Queen conch reproductive stocks in the central Bahamas: distnbution and probable sources. Fish. Bull. U.S. 92:171-179. Stoner, A. W. & J. M. Waite. 1990. Distribution and behavior of queen conch, Strombus gigas. relative to seagrass standing crop. Fish. Bull. U.S. 88:573-585. Sutherland, J. P. 1987. Recmitment limitation in a tropical intertidal bar- nacle: Tetraclita panamensis (Pilsbry) on the Pacific coast of Costa Rica. J. Exp. Mar. Biol. Ecol. 113:267-282. 420 Stoner et al. Victor. B. C. 1986. Larval settlement and juvenile mortality in a recruit- ment-limited coral reef fish population. Ecol. Moiiogr. 56:145-160. Weil. M. E. & R. Laughlin. 1984. Biology, population dynamics, and reproduction of the queen conch, Strombus gigas Linne, in the Archi- pielago de Los Roques National Park. J. Shellfish Res. 4:45-62. Wethey. D. S. 1984. Spatial pattern in barnacle settlement: day to day changes during the settlement season. J. Mar. Biol. Assoc. U.K. 64: 687-698. Yeung, C. & M. F. McGowan, 1991 . Differences in inshore-offshore and vertical distnbution of phyllosoma larvae of Panulinis, Scyllarus and Scvliarides in the Florida Keys in May-June. 1989. Bull Mar. Sci. 49:699-714. Yoshioka. P. M. 1982. Role of planktonic and benthic factors in the pop- ulation dynamics of the bryozoan Membranipora memhranacea. Ecol- ogy 63:457-168. Yund, P. O . S. D. Gaines & M. D. Bertness. 1991. Cylindncal tube traps for larval sampling. Limnol. Oceanogr. 36:1167-1177. Journal of Shellfish Research. Vol. 15. No. 2. 421^32, 1996. THE MANGROVE SNAIL THAIS KIOSQUIFORMIS DUCLOS: A CASE OF LIFE HISTORY ADAPTATION TO AN EXTREME ENVIRONMENT VOLKER KOCH AND MATTHIAS WOLFF Ceiiicr for Tropical Marine Ecology Klagenfurter Sir. Gehaiide GEO 28233 Bremen. Germany ABSTRACT This article describes Ihc population ecology of Tluiis kinsquifonnis Duclos, the dominant predatoi^ gastropod of the root system of Costa Rican mangroves. T. kiosquiformis was shown to cope with the extreme living conditions of its habitat (risk of desiccation and overheatmg through several hours of daily air and sun exposure, strong salinity, and current changes during the tidal cycle) by using the following strategies: ( 1 ) extremely slow growth (~ I mm/y). cessation of growth at the onset of maturity (at —24 mm in shell length), (2) maintaining high interindividual plasticity in growth and shell thickness as a response to abiotical conditions, food availability, and population density, and (3) migrating ontogenetically for the benefit of lowering desiccation and predation mortality. Because of its high density and biomass (192.2 ± 102.4 g wet weight/m"), and the predation pressure it exerts mostly on the barnacles of the mangrove roots, T. kiosquiformis seems to occupy a central role in maintaining the functioning and productivity of mangroves through "cleaning" their root system from the encrusting fauna. KEY WORDS: Costa Rica, mangroves, population ecology, gastropods INTRODUCTION Thaidid gastropods evolved in the early Miocene (Vemieij 1978) and are distributed worldwide from tropical to boreal areas, indicating the large adaptive potential of this family. They occur in intertidal and subtidal shallow waters, often as a dominant inver- tebrate predator within their habitats (Mengc 1978). Studies on the ecology of thaidid gastropods revealed a complex set of mecha- nisms to deal with the harsh environmental conditions they en- counter in the intertidal habitat. These include prey type selection motivated by energetic considerations (Palmer 1984); prey size choice to minimize the risk of dislodging at wave-exposed sites (Richardson and Brown 1990); optimized foraging behaviour in relation to prey abundance and mortality factors such as predators and desiccation (Menge 1978, Spight 1983, Fairweather 1988); switching to anaerobic metabolism by aperture closing to avoid desiccation (Cantera et al. 1980); size-dependent zonation related to factors like food, predators, shelter, and desiccation (Butler 1979); and modification of shell thickness as a response to differ- ent predation pressures by crabs (Kitching 1977, Palmer 1985, Geller 1990). Menge (1978) concluded from his study on preda- tion exerted by Thais lapillus on a rocky shore that each snail has to be considered an individual, where life history traits and phe- notype are as important as extrinsic factors (e.g., actual habitat conditions). West (1988) also reported high interindividual varia- tion in prey preferences and growth of Thais melones on the Pa- cific Coast of Panama. Great individual plasticity in behavioural, physiological, and morphological responses to a harsh environ- ment seems thus to explain the success of this family. Thais kiosquiformis Duclos is among the most abundant snail species and probably the most important invertebrate predator in the mangrove swamps of the upper Gulf of Nicoya on the Pacific Coast of Costa Rica (10°N, 85°W). It is distributed along the Pacific and Atlantic Coasts from Baja California to Peru in the intertidal at salinities of 5-30%r, living mainly on mangrove roots, rocks, and rotting trunks (Keen 1971. Cantera et al. 1980). It is strictly carnivorous, preying on balanids, bivalves, and other gas- tropods by drilling a hole in the prey's shell or by introducing its foot. Cannibalism does not seem to occur. Specimens can survive desiccation up to 9 d by closing the aperture tightly and switching to anaerobic metabolism. The species grows to approximately 50- 55 mm in shell length; the strong shell is muddy brownish with well-developed spines. Maturity is reached at 24-32 mm shell length (Cantera et al. 1980). The objectives of this study were ( 1 ) to describe the population structure (density, biomass, horizontal size distribution) of 7". kios- quiformis at study sites differing in freshwater influence, sediment characteristics, height above low water level, and total mangrove extension; (2) to estimate growth and mortality rates at the differ- ent study sites; and (3) to estimate food consumption and identify main prey items. The overall goal of the study was to explain the mechanisms that not only promote survival under the extreme conditions of the mangrove habitat but also allow this species to maintain high abundance and biomass in the mangrove root com- munity. MATERIALS AND METHODS Study Area and Sampling Sites The Gulf of Nicoya is a tectonic estuarine embayment located on the Pacific Coast of Costa Rica. Central America (Fig. I ). The gulf is divided into a shallow upper part ( —20 m) and a lower part ( >200 m in depth), which is bordered by the Puntarenas Peninsula in the east and San Lucas Island in the west. Because of seasonal upwelling events, this region is the most productive fishery ground in Costa Rica. Seasonality on the Central Pacific Coast of Costa Rica is very pronounced, with 89% of the annual precipitation occurring from May to October. Average total precipitation for Puntarenas is ap- proximately 1,550 mm/y (Thomas 1988). The annual mean air temperature for Puntarenas is 27.35°C (Janzen 1991); water tem- peratures in the upper gulf range from 28 to 30°C. The water body of the upper gulf is heavily influenced by freshwater and sediment input of the Rio Tempisque, located at the northern end, and by numerous smaller rivers, especially along the eastern shores (Peterson I960). The tides are semidiurnal (12.4 h); amplitudes vary between 1.8 and 2.8 m with a mean of 2.3 m (Peterson 1960. Voorhis et al. 1983). During the rainy season, surface salinity in the upper gulf may drop down to 5'?f near river mouths during low 421 422 Koch and Wolff Figure 1. Gulf of Nicoya with tlie tliree study sites (solid circles) — Punta (PTA) Morales (P), Cocoroca (C), and Jicaral (J). Population Ecology of the Mangrove Snail Thais KiosQuiroRMis 423 tide (this study). The outer coastline of the upper gulf is eovercd with extensive mangrove stands, dominated by red mangroves (Rhizophoni mangle and Rliizuphiini hanisimi). Two sampling sites were chosen in each of the three estuaries of Punta Morales (PI, P2), Cocoroca (C4. C5). and Jicaral (J6. J7) for a representative cross-section of the upper gulf (Fig. 1). The vegetation consisted exclusnely of the two Rhnophoni species mentioned above. Large mudflats extended seaward at each loca- tion. Freshwater input was highest at Cocoroca (Rio Lagartol. followed by the Rio Jicaral and the Rio Quebrada Grande in Punta Morales (Fig. 2). The latter only carries freshwater during the rainy season, and even then, the input is very low (Gocke et al. 1981, this study). The increase in salinity in November marks the end of the rainy season. At high tide, the water level at the sam- pling sites of C and J was approximately 1-1.5 m; in P. it was 1.5-2 m above the sediment, coinciding with the upper limit where barnacles occurred. The sediment in Punta Morales was coarser than that at the other sampling sites, with stones and rocks contrary to the uniform sandy/muddy Cocoroca and Jicaral sites. Accordingly the "sink-in depth"' varied considerably: J > C > P. Sampliiif; Strategy Twelve length frequency samples were taken at about monthly intervals along transects at PI, P2, and C4 from September 1993 to July 1994. In C5, J6, and J7, only 10 samples were taken from October 1993 to July 1994. An additional sample of tagged snails was taken in February 1995 in PI . Surface water salinity and water temperature were measured at the adjacent water line (refractom- eter, ±0.59?(; mercury thermometer. ±0.2°C). Each transect con- sisted of 10 squares of 60 • 60 cm each (0.36 nr), placed 1 ni apart from centre to centre, with a 10-m line tagged at 1-m inter- vals. This line was positioned vertical to the mangrove edge (Fig, 3). Because a solid frame could not be used because of the man- grove roots, an inch rule was taken instead. All T. kiosquijonnis found within the square were taken to the laboratory. Total shell length was measured to the nearest 0.5 mm with a caliper. Snails from each square were measured separately to determine length frequency patterns along the transect. After measuring, specimens were returned to the transect to avoid introducing artificial mor- tality into the population. This was not done in Jicaral because of the large distance to the laboratory (1 h by boat). On the first sampling dale, a different strategy was used to determine the appropriate square size: in addition to the basic design, each square was divided into four subquadrats of 30 • 30 cm each (0.09 m"). Mean length, standard deviation, and variance of the subsamples were calculated to determine whether the vari- ance changed as a function of sample area. The 60 ■ 60 cm quad- rat size gave homogeneous results even for densities <35 snails/ m", whereas the smaller quadrat size (30 • 30 cm) did not perform well under low densities, where number of snails, mean length, and standard deviation differed significantly among the subsam- ples. Therefore, the larger quadrat size was selected for subse- quent sampling during the study (length frequency sampling and biomass determination). GroKih and Mortality The growth of T. kiosqiujonnis was determined from the length frequencies by use of the seasonalized von Bertalanffy growth equation (Pauly and Gaschiitz 1979) as implemented in the FiSAT program (Gayanilo et al. 1994): U(l -(K(l-t„l + CK/2-rr(l-U) where L^. is the asymptotic length, L, is the length at time t, K is the growth constant. t„ is the age at length zero, C lies between 0 and 1 and describes the growth amplitude (of a sine function), and t., is the starling point of the growth oscillation. The program restructures the data by calculating a running average over five length classes, dividing each length class by this value and subtracting 1 from the result, which creates peaks and troughs. The program traces growth curves with different sets of parameters through the length frequencies and selects the curve that hits as many peaks and avoids as many troughs as possible. 35 r 30 25 ^ 20 '5 ■■s '5 10 5,9.93 26.9.93 18.10.93 3 1 1 93 27 M 93 28 12.93 26.1.94 5.3.94 sampling date Figure 2. Surface salinity at the study sites P, C, and J durin)> low tide. 424 Koch and Wolff Sampling strategy mangrov* edge VI VII VII 1 m Figure 3. Sampling strategy. Letters (l-X) marl^ quadrat numbers. The length measurements of each transect were grouped into 1-mm intervals. Because of the striking differences in biomass and length distributions between the transects, the von Bertalanffy pa- rameters were calculated separately for each transect. Because the analysis of the whole range of length frequencies did not give satisfactory results (goodness of fit. Rn < 0.2 in all cases), the youngest visible cohort was determined by eye and reanalysed separately (PI and P2, =£23 mm; C4, s:22 mm; C5 and J6, ^25 mm), as described in Wolff (1985). In J7, growth was not deter- mined because of the low number of snails found during each sampling date (<30). The growth performance index (0'). which allows for the com- parison of growth between different species (Pauly and Munro 1984), was then calculated with the growth parameters K and L^: 0' = log,o K + 2 ■ logio L^ As a second method to estimate growth parameters and to compare with the results of the length frequency analysis, 643 snails were marked and monitored over the sampling period. A paint marker was used to mark 230 snails; 186 were marked with quick-drying oil paint. The numbers were written on the shell with waterproof markers. Because both methods failed, a sandwich technique was used, applying two layers of nail paint on the cleaned shell near the aperture. The numbers were written with india ink, and the mark was sealed with "crazy glue." This method was applied to 413 snails, of which 186 specimens had old marks (re-marked). The animals were released near the sampling site PI . On each sampling occasion, a search of about I h was conducted for marked snails, which after collection, were measured with a caliper to the nearest 0.1 mm and subsequently released at the same spot. Estimates of von Bertalanffy growth parameters were derived by the use of the Munro plot (Munro 1982). K„, ln(U Li) - ln(L:.: - L:) where L,, L., t,. and t^ are the lengths and times of marking and recapture, respectively, L, is as described above, and K,,^ _ ^^, is the growth constant over the time interval. The model allows for calculating the growth constant K for each individual. The sample from February 1995 was treated separately because the transect had been left undisturbed for nearly I y, avoiding recapture stress and habitat disturbance through frequent sampling. A linear re- gression of maximum shell length {L^^J versus average density at the transects was calculated, expecting a negative correlation of Lmax with density. L^^^ was defined as the mean of the largest 3% of the snails at the respective transect (Pauly 1984). Total mortal- ity (Z) was estimated with the mark-recapture data by use of a formula proposed by Gulland (1969): In N^ = a -I- b • r' where N^ is the number of recoveries per time interval r', a is the y-intercept, and b is the slope of the regression, which provides the estimate of Z (with sign changed). This method can only be ap- plied to data where the marking procedure is performed at one time (e.g., a few days), the sampling effort is roughly similar, and mark shedding docs not occur (Pauly 1984). Therefore, the October 1993 census was taken as the starting point. Only snails marked or recaptured (and re-marked) at this date were included in the anal- ysis. This ensured that only specimens with sealed marks (no mark shedding) were used for mortality estimation. The February 1995 recapture data had to be corrected for sampling effort because the Population Ecology of the Mangrove Snail Thais kiosquiformis 425 area was searched for 4 h. exceeding normal effort by a factor of four. Population Structure, Biomass, and Density For each quadrat, the median length was calculated (this was preferred to the mean length because of the lower sensitivity to "outliers") with the pooled data of all samples taken from Sep- tember to March to prove whether different length groups prefer distinct zones along the 10-m transect from the mangrove border inwards. In addition, the proportion of juveniles (ss20 mml of the total was calculated for each quadrat to see if they prefer a distinct zone within each transect and to compare the dominance of juve- niles between the transects. Total shell length was converted to dry weight with 87 speci- mens of T. kiosquiformis . To do so. the shell was broken, and the cleaned tissue was placed m aluminum dishes and dried for 100 h to constant dry weight at 65°C. Dry weight was determined with a precision of 0. 1 mg. A potential regression of the form y = a ■ x*' was used (Table 1 ). The last four samples were excluded from the size distribution and from the biomass calculations because of possible sampling bias because they were not taken by the authors. Dry weights of all individuals of a length class were calculated from the pooled length frequency data with the regression given above. Biomass values of the length classes were summed and divided by the total area (in square meters) sampled, yielding an average biomass of T. kiosquiformis for the upper gulf. The bio- mass of each sampling quadrat was calculated by pooling the length frequencies for the respective quadrat. Food Intake The in situ food intake of three size groups of T. kiosquiformis (18-22 mm; 23-27 mm; 28-32 mm) was estimated over a period of 14 d in the field (Punta Morales). Ten snails of each group were placed in a mesh wire cage (height, 1 m; 0, 40 cm), situated approx. 0.5 m above mean low water level at the mangrove edge. The 10 snails were weighed collectively before and after the ex- periment to the nearest 0. 1 g. Wet weight again was converted to flesh weight gain by use of a linear regression of wet weight versus flesh weight (Table 1). Wet weights of each prey species were also registered before the experiment (for balanids that were offered on pieces of aerial roots, percent area covered was estimated). The following eight species (corresponding to the most abundant po- tential prey in the area) were placed in each of the three cages: Balanus sp. (Crustacea). Litloraria varia. Littoraria fasciaia. Anachis, rugosa. Cerilhium slercusmuscarum (Gastropoda), Brachidonlis. puntarensis. Pinctada mazallanlica (juv.), and Cardita affuus (Bivalvia). P. mazantlantica is not common in mangrove forests, but a large spatfall had occurred on the man- TABLE 1. T. kiosquiformis: Regression Parameters of Dry Weight Versus Shell Length and Flesh Weight Versus Wet Weight I'sed for Biomass Estimation and Calculation of Daily Ration. Total Length/Dry Weight Wet Weight/Flesh Weight a = O.OOOIO b = 3.187 r^ = 0.976 n = 87 a = -0.00700 b = 0.129 r^ = 0.937 n = 87 grove roots shortly before the experiment started, so this bivalve was included. Two months later, this species had disappeared from the area. After the experiment, the surviving prey organisms were counted and weighed together with the empty shells. Because all prey had shells, weight differences measured after the experi- ment consisted only of flesh weight consumed by the predator. Daily food intake was expressed as % BWD (flesh weight con- sumed daily/wet (flesh) weight of predator). The weight increment of r. kiosquiformis was expressed as % wet weight gain/14 d. Prey preference was not determined; the relative abundance of prey species and their availability in the experiment were not compa- rable to normal habitat conditions. Some prey specimens (8 of 156) were not found after the experiment and had to be excluded as not eaten. RESULTS Growth and Mortality The length frequency analysis yielded similar parameters for the different sampling sites (Fig. 4), where the ranges are: K, 0.18-0.2; L.,., 48-50 mm; C, 0.575-0.65; Wp, 0.775-1.0. The goodness of fit (R„) of the growth curves ranged between 0. 171 (C5) and 0.341 (J6). The average growth performance (0') was 2.66. with values for K and L, of 0. 19 and 49 mm, respectively. The tagging experiment conducted in Punta Morales yielded results that differed from those of the length frequency analysis. Recapture data from September 1993 to March 1994 and from March 1994 to February 1995 are presented separately, because the latter consisted of individuals that were left undisturbed for 1 y (Fig. 5) The points can be divided roughly into three groups; specimens with positive K values, >0.02 (I); others with K values around zero. 0.02 > K > -0.02 (II); and snails with negative K values. <-0.02 (III). Between September and March, only six juvenile and subadult snails (<24 mm) showed positive K values. Most specimens did not grow at all; snails larger than 25 mm tended to show negative K values. For the period from March to February, the smaller animals had the highest K values, decreasing more or less linearly to zero at a shell length of 24 mm. The same division of points was applied here, but because no negative K values occurred (and no specimens >26 mm), only the first two groups are represented. Mean density at the transects was nega- tively correlated (R- = 0.944) with maximum shell length (Fig. 6). Highest L„^^ values occurred at J7 and J6. where density (and biomass) were lowest, whereas high densities (PI and C4) related to low L„3.j values. P2 was excluded from the calculation because the population at this transect consisted almost exclusively of ju- veniles and subadults. The estimation of the total mortality rate (Z) from the tagging data (Fig. 7) yielded a value of 0.178/y (c.i. 0.132-0.225). The resulting fit of the regression was good (R" = 0.951; p = 0.0002). Population Structure, Biomass, and Density The distribution of length classes along the transects (Fig. 8) shows a clear prevalence of smaller animals in the inner zone. The median length decreases from the mangrove border to the interior part, stabilizing 5-8 m inwards from the mangrove edge. The graph also demonstrates the large differences in median shell length between the transects. The subpopulation in J7 consisted of the largest snails, followed by J6, C5. C4, PI, and P2. In P2. the 426 Koch and Wolff 35 n=471 n^^Sh n=^XI n^H n=4% n=4M n=33-l n=ly7 n^U^ n^2l4 n=\M n=H\ ^ n= 114 n=2m n=2»7 n=ll4 n=14n n=l(* n=266 n=l7l n=l4X n=n4 n=ll| n=ll7 3$ • 30 ' PI Loo = 48nim K =018 C = 0.575 Wp = 0.9 Rn =0.287 P2 Loo = 49mm K =0.18 C =0.6 Wp = 0.775 Rn =0.211 E S ^ JS Hi 30 e Ji S 20 4> n=^«. n=V12 n=242 n=2^^ 11=28^ n=2«> n=276 n=22(. „=2I>. n=lMl n=21'» n=l66 C4 Loo = 48mm mWri-i-H n=2M n=V<(. n=2S2 n=262 n=2J4 n=J)4 n=276 n=126 n=187 n=162 ii444^^-^-^i- I I I I I ^^^-^ I I 35 . 3* ' 25 • » ■ 15 ' mill n=l(« n 141 n-125 „=|.n ,,= 118 n=l(r; n=87 n=<»2 n=% mm I I I I I I I I I I I SONDJ FMAMJ J K =0.2 C =0.6 Wp = 0.95 Rn =0.323 C5 Loo = 50mm K =0.2 C =0.6 Wp = 0.9 Rn =0171 J6 Loo =49mm K =0.2 C = 0.65 Wp = 1.0 Rn =0.341 tiine (months) Figures. T. kiosqiiiformis: growth at the five transects, calculated from the length frequencies with ELEFAN. L^ and K von Bertalanffy growth parameters; C, constant of growth oscillation; Wp, winterpoint; Rn, goodness of fit. Population Ecology of the Mangrove Snail Thais kiosquiformis All 0,15 , 0,1 — 0,05 a) -0,05 -0,1 - o oo o o O 00 o o o - COD oo OOOdS oo (QDCO) OCOQID OflD O dnSOO (IE>0 QD CO O CBD O Om oo O O OG& QD <]]) o o o O OOO oo OOOOD (EDO O O -0,15 16 18 20 22 24 26 28 30 32 shell length (mm) 0.1 b) • • ♦ group 1 0,08 r • * • • <-' group II 0,06 p 1 i 0,04 - * ♦ 0,02 - O O o 0 1 1 1 1 CB O 0 O O 0 i 14 16 18 20 22 24 26 shell length (mm) Figure 5. 7", kiosquiformis: growth constant K as related to shell length for individuals of the marking experiment, (a) Sampling period from September 1993 to March 1994. (b) Sample taken in February 1995. Group 1, K > 0.02; group II, 0.02 < K < -0.02; group III, K < - 0.02. decrease in median length was not very pronounced, but because it was the transect with the smallest snails, large changes in size frequency distribution could not be expected. The dominance of juveniles =£20 mm is therefore clearly strongest in P2. followed by PI and C4 (Fig. 9). At the other transects, only O-lO^f juveniles were found. At the first three transects, the dominance of juveniles increased inwards from the mangrove edge, indicating a gradient along which snails of different age classes are distributed. Highest density values (ind./m"), obtained at the transects, were: 222 at PI. 139 at P2. 194 at C4, 111 at C5, 56 at J6. and 19 at J7. Biomass. averaged over transects and sampling period, was highest in quadrats III-V (2-A m from the mangrove border) (Fig. 10). The high standard errors are partly explained by the high variation of biomass values between transects, differing by a factor of up to 6. Average biomass and standard deviation of T. kiosqui- formis, calculated over transects and sampling period, were 6.37 ± 3.41 g dry weight/m" (192.2 ± 102.4 g wet weight/m"). Food Intake The average daily ration was lowest in the medium group (1.3% BWD wet weight; 10.7% BWD flesh weight), followed by the smallest group (1.6% BWD wet weight; 13% BWD flesh weight) (Table 2). The largest animals had the highest daily ration (2.0% BWD wet weight; 16% BWD flesh weight). Liuorina spp. accounted for nearly 70% of prey eaten in the experiments, fol- lowed by the pearl oyster P. mazattanlicu with 19% and the small mytilid B. puntarensis with 10%, Under natural conditions, how- ever, T. kiosquiformis primarily feeds on balanids (pers. obs.). The results of the fleld experiment (Table 2) show that only the smallest size group had slight positive growth in wet weight (0.6%/ 1 4 d), whereas the larger groups (23-27 and 28-32 mm) lost weight during the experiment (-1.9 and - 1.3%/ 1 4 d, re- spectively). DISCUSSION Growth and Mortality The length frequency analysis yielded very low K values (0.18-0.2) for T. kiosquiformis for all transects when compared with the reports of Cantera et al. (1980), who gave growth rates of approximately 0.5 mm/mo (K/y = 0,29) for this species on the Colombian Pacific Coast. Although the goodness of fit (Rn) was rather low for our K estimates (0.171-0.341), the calculated pa- rameters are very similar between transects and seem reliable. The growth performance index (0' = 2.66) is correspondingly low and seems more comparable to that of boreal than of tropical gastropods, which normally exhibit values between 3.2 and 4.7. Only two boreal thaidids and some Liuorina species are reported to have similar low values (Wolff 1994). The results of the marking experiment indicate still much lower growth rates than were found with the length frequency analysis. Of nearly 300 recaptured snails, only a few showed positive growth during the sampling period. One might suspect marking and/or recapture stress to be the cause, but this does not seem probable for the following reasons: (1) marking was conducted by three different methods (see Materials and Methods); (2) marking procedure and recapture never lasted longer than 2 h, whereas snails are naturally exposed for up to 5 h during low tide; (3) many marked snails were recaptured the first time alive after 3-5 mo without showing any measurable growth; and (4) the sample taken in February 1995 yielded basically the same results, although the transect had been left undisturbed for nearly 1 y. The decrease in total shell length in some specimens is due to shell erosion at the apex, where the periostracum can be damaged, especially in older animals, leaving the calcareous shell structures in this part uncov- ered. The interindividual variability in growth — seen in the wide scatter of individual K values — is very high, but several authors report similar variabilities for other molluscs (e.g., Moore 1972, Broom 1982, Sainsbury 1982. Wolff 1987). This variability can probably be explained through small-scale differences in habitat structure but possibly also by genotypical differences between specimens (Wolff 1987). Unexpectedly, K, generally considered to be constant for a species over its whole life cycle, clearly decreased with size, and it seemed reasonable to divide the individual K values into three groups: positive values for snails =s24 mm, values around 0 (oc- curring in the whole size range), and negative K values for snails >24 mm [one exception]. It thus seemed that 24 mm, which marks the lower limit for the onset of maturity (Cantera et al. 1980), can be taken as the turning point for the snails in PI , where growth ceased completely. This suggests ontogenetically based metabolic changes related to the onset of matunty. This pattern can be described for the study site PI . where population densities were highest. However, for the other study sites, L^^^ values were 428 Koch and Wolff E E 45 40 35 30 25 • EJ7 y =43.1 19- 0.3027 'x r^ = 0.9435 EJ6 EC 5 EC 4 J'M ■ PM2 excluded 10 15 20 25 30 35 40 mean density / 0.36 m Figure 6. T. kiosquiformis: maximum shell length (L„,^ ) vs. average density at the six transects. P2 was excluded from the fitted regression (see text). significantly liiglier, which suggests that growth in these areas continues in older individuals. Thus, it might be speculated that somatic growth ceases early under high population densities (re- sulting in smaller maximum length) in response to limiting food supply. Under these conditions, energy may be used exclusively for gonad formation and maintenance metabolism. A negative ef- fect of high densities on growth and maximum length through intraspecific competition for microhabitats and food resources has also been reported for other intertidal gastropods (e.g.. Under- wood 1978). This phenomenon could naturally not be detected by the length frequency analysis because only juvenile specimens, which were still growing, were used. The thick shell of T. kiosquiformis. being the most robust among the snails occurring at Punta Morales (Borjesson and Szeli- towski 1989). is energetically costly (Palmer 1985. Kitching 1986) and might be another factor explaining the very slow growth rate of 7. kiosquiformis. In related species, thick-shelled morphs were also shown to have a decreased tissue growth (Thais lamellosa. 5 4,5 • -? Z C-i = 0 951 = 0178 = 0 132-0 225 4 - • -^ 3.5 ' • ^\^ 3 - 2,5 - 2 1 1 1 I -\, 0 4 8 12 16 time (month!) Figure 7. T. kiosquiformis: total mortality rate (Z) as derived from Gulland's (1969) method. Palmer 1981); slower shell growth (Purpura species. Wellington and Kuris 198.'?); higher food intake as the result of increased production, maintenance, and locomotion costs (T. laptllus and T. lamellosa. Palmer 1992) and less offspring production (Thais emarginala. Geller 1990) when compared with thin-shelled morphs. Finally, energy costs for adaptations to environmental stress, which is supposed to be very high in tropical intertidal areas (Moore 1972. Vermeij 1978. Garrity 1984). may contribute a significant part to the total energy budget, further reducing energy available for growth (Russell-Hunter 1985). We believe that a combination of the above-mentioned factors is responsible for the extremely slow growth of T. kiosquiformis. The results of the length frequency analysis possibly reflect the growth potential of this species when conditions are favourable and food supply is not limiting. In J6. this may have been the case, because density was low, animals grew near to their asymptotic size, and goodness of fit (Rn) for the resulting growth curve was very good. If high population density and biomass are maintained while individual growth is very slow, survival must be maximized. Our mortality estimate (Z = 0.178), which is at the low end of the range reported for tropical and subtropical gastropods (0.1-1.66) (Sainsbury 1982. Appeldoom 1987. Appeldoom 1988. Prince et al. 1988. Wolff 1989. Debrot 1990). is an indication thereof. This low mortality is probably due to; ( 1 ) the strong antipredatory char- acteristics of the shell (thick shell, narrow aperture, stout spines), which reduce predation pressure by crabs and fish (Vermeij 1978, Palmer 1979, Palmer, 1985, Bertness and Cunningham 1981. Wellington and Kuris 1983); (2) the high tolerance against desic- cation (Cantera et al. 1980); and (3) the ability to minimize energy expenditure (extremely slow growth, growth inhibition at the onset of maturity), using a "sit and wait"" strategy. These results con- tradict Alongi's (1989) general assumption that turnover rates as well as predation mortalities in the tropics are higher than in tem- perate latitudes. The method used may even have overestimated mortality, be- Population Ecology of the Mangrove Snail Thais kiosquiformis 429 30 22 - 20 P1 J L J L 35 C4 15 I ^ ^ L. I I I I L II III IV VI VII VIII IX 40 35 - 30 25 20 III IV VI VII VIII IX Quadrat No. Quadrat No. Figure 8. T. kiosquiformis: box plots of the size distributions along the six transects, with the combined data of all sampling dates. Horizontal line within the box indicates the median, first upper and lower quartiles are given by the vertical edges of the box, vertical bars indicate the whole data range, and open circles are extreme outliers, which were excluded from the calculation. cause mark shedding may have occurred to a small extent and because snails could have migrated out of the area. In addition, specimens well hidden might not have been found. Population Structure, Biomass, and Density The observed distributional pattern — with the juveniles pre- dominating mangrove inwards and the larger specimens towards the mangrove edge — probably results from the active distribution of the respective size groups along a gradient of prcdation and desiccation, as described for several gastropods (e.g., Vermeij 1972. Butler 1979. Garrity 1984). The most important gastropod predators in the area are puffer fish iSphoeroides species and Di- odon species) and rays, which were frequently observed foraging in Punta Morales (Jerome 1987, Whitey 1990. pers. obs.). Pre- dation pressure is probably more severe at the mangrove border than inside the forest for the following reasons: (1) the root system is much denser inside, making foraging more difficult here: (2) foraging time is reduced inside the forest because of shorter tidal 430 Koch and Wolff I II III IV V VI VII VIII IX X Quadrat no. Figure 9. 7". kiosquiformis: ratio of juveniles (=520 mm) to total num- ber of specimens ± standard errors along the six transects. inundation (higher elevation of the sediment); and (3) hght con- ditions are better at the mangrove border (less shading through the canopy), making visual recognition of prey specimens easier. Be- cause large T. kiosquiformis are less susceptible or even immune to predation, as shown by Palmer (1979). they can live at the mangrove border, where barnacles are more abundant than in the inner zone (pers. obs.). whereas smaller snails are largely re- stricted to the inner forest, where predation is less intense. Irradiation is more intense at the mangrove border, where the canopy is less dense than mangrove inwards, resulting in relatively higher temperature and desiccation. Wind is also stronger at the border because air movements are dampened by roots and leaves, reducing desiccation stress in the inner zone. Generally, larger individuals of a given gastropod species can tolerate desiccation much better than juveniles (Moore 1972, Vermeij 1972. Vemieij 1978, Underwood 1978). which allows them to stay further out- side than juveniles. Personal observations confirmed the migration pattern of smaller specimens towards the inner zone, whereas large I II III IV V VI VII VIII IX X Quadrat no. Figure 10. T. kiosquiformis: average biomass distribution ± standard errors for each quadrat along the transect. Data from all transects combined. animals dispersed in all directions or stayed at the border (Koch unpubl. obs.). The biomass distribution along the transect, with its maximum 2-A m from the mangrove border, can be explained by the above- mentioned mechanisms affecting the distribution of respective size groups. Middle-sized snails, which account for the largest portion of total population biomass, were found in highest densities 2—4 m from the mangrove border, where stress is already reduced but food supply is still high. Further inwards, where food seems less abundant but where stress is minimized, the smallest snails occur and survive the most vulnerable life stages. The average biomass of T. kiosquiformis in the Gulf of Nicoya of 6.37 ± 3.41 g dry weight/m-( 192.2 ± 102.2 g wet weight/m') is very high when compared with values of other mangrove areas. Lalana Rueda and Gosselck ( 1986) found values between 8 and 17 g dry weight/m" for the whole epifauna during the rainy season, with much lower biomasses in the dry season. In Taiwanese man- groves, biomass values of 131—406 g wet weight/m~ are reported for the whole epifauna (Wu et al. 1992). These values are similar to that of T. kiosquiformis and reflect the high secondary produc- tivity of the mangrove community in the study area. In terms of biomass and abundance. T. kiosquiformis is the most important species in the mangrove root community of the study area. Food Intake The overall estimate for the average food intake of 1% BWD wet weight (9% BWD flesh weight) is similar to that of the related species. Thais canmfera (10% BWD tlesh weight and to that of TABLE 2. T. kiosquiformis: Food Intake, Percentage Consumed of Each Prey, and Weight Increment of the Three Size Groups During the Field Experiment |I4 d). Size Groups (mm) Parameter 18-22 23-27 28-32 (n = 10) (n = 10) (n = 10) Starting weight (wet) of 10 T. kiosquiformis (g) Final weight (wet) of 10 T. kiosquiformis (g) Mean Oesh weight of 10 T. kiosquiformis (g) during the experiment ( 14 d) Total food intake of 10 T. kiosquiformis (g) Daily ration in % wet weight/day Daily ration in % flesh weight/day Weight contribution of each prey species in % L. varia L . fasciata B. pumarensis Pinclada sp. C. affinis A. rugosa C. stercusmuscarum Balanids Total weight gain of T. liiosquiformislXA d (%) 14.71 32.02 44.43 14.8 31.4 43.84 1.83 4.02 5.62 3.33 6 12.61 1.6 1.3 2 13 10.7 16 0 2 13 56 78 54 16 3 11 28 17 12 0 0 10 0 0 0 0 0 0 0 0 0 0.6 -1.9 1.3 Population Ecology of the Mangrove Snail Thais kiosquiformis 431 the naticid snail Natica maculosa (7.57( BWD flesh weight), which occurs in tropical mudflats (Broom 1982). By the use of average biomass and daily ration, the population of T. kiosqui- formis consumes approx. 2.5 g tlesh weight/m' per day. We do not have any reasonable explanation for the differences in daily rations (%BWD) found between the three size groups, with the largest specimens having the highest food intake, followed by the smallest and the middle sized. The strong preference for littorinid snails in the experiment is most probably not representative of field conditions, where T. kiosquiformis was found to feed primarily on barnacles (Cantera et al. 1980. Perry 1988. pers. obs.). Littorinids are not readily avail- able as prey because they occur much higher in the mangroves of Punta Morales (Jerome 1987. Withey 1990) and escape when T. kiosquiformis is present (pers. obs.). a behaviour also reported for other littorinid snails iMcKillup 1982. Fairweather et al. 1984). Furthermore, thaidid and muricid snails are reported to be capable of choosing the most profitable prey (in terms of energy per unit of effort) and readily switch their preferences when a prey species yielding higher profit is made available (Bayliss 1982. Palmer 1984. Carroll and Wethey 1990). The weight changes during the experiment seem to confirm the results of the growth analysis, because only the smallest specimens gained weight, whereas the larger snails lost weight during the experimental period. The growth of mangrove trees in the study area is possibly highly dependent on the density and biomass of T. kiosquiformis. Perry (1988) reported a 50% growth reduction in the aerial roots of R. mangle when covered with barnacles. She found T. kiosquiformis and hermit crabs to be the most important predators, but the latter occurred in much lower densities along the transects studied (Koch unpubl. obs.). k similar situation was found in Belize, where mangrove growth also depended on bar- nacle coverage (Ellison and Farnsworth 1992). The dominant predator in that system was the snail Melongena melongena. Our study thus suggests that T. kiosquiformis structures the mangrove root community in the study area through the predation pressure it exerts on the community and maintains or enhances the produc- tivity of the mangrove trees by constantly cleaning the root system of its encrusting epifauna. ACKNOWLEDGMENTS This study was undertaken as part of a Masters' Degree and was financially supported by the "Studienstiftung des Deutschen Volkes." We thank the Costa Rican friends and colleagues who helped with the practical work and the logistics, especially Roci'o Cordoba. Sergio "Cuhado'" Segura. and Jose A. Vargas. LITERATURE CITED Alongi. D. M. 1989. Ecology of topical soft-bottom benthos: a review with emphasis on emerging concepts. Rev. Biol. Trap. 37(1);85~100. Appeldoora. R. S. 1987. Assessment of mortality in an offshore popula- tion of queen conch. Sirombus gigas L . in Southwest Puerto Rico. Fish. Bull. 85(4);797-804. Appeldoom. R, S. 1988. Age dctemiinallon. growth, mortality and age of first reproduction in adult queen conch, Strombiis )>iga.<; L.. off Puerto Rico. Fish. Res. 6:363-378. Bayliss. D. E. 1982. Switching by Lepsiella vinosa (Gastropoda) in South Australian mangroves. Oecologia 54:221-226, Bertness. M. D. & C. Cunningham. 1981. Crab shell-crushing, predation and gastropod architectural defense. J. Exp. Mar. Biol. Ecol. 50:213-230. Borjesson. D. L. & W. Szelitowski. 1989. Shell utilization and predation in the hermit crab Clihanarius paiuimensts. Stimpson, in a tropical mangrove estuary. J. Exp. Mar. Biol. Ecol. 133:213-228. Broom. M. J. 1982. Size-selection, consumption rates and growth of the gastropods Natica maculosa ( Lamarck ) and Thais curinifera ( Lamarck ) preying on the bivalve Anadara granosa. J . E.xp. Mar. Biol. Ecol. 56:213-233. Butler. A. J. 1979. Relationships between height on the shore and size distributions oiThais spp. (Gastropoda: Muncidael. J. E.xp. Mar. Biol. Ecol. 41:163-194. Cantera. J. R.. R. Contreras. F, Bon-ero. E. Buttkus & F. Zapata. 1980, Historia natural del gastropodo del mangle Thais kiosquiformis Duclos. 1832. en la costa del Pacifico Colombia. Memorias del seminario organizado por UNE SCO (Call. 27.1 1-01. 12. 1978): 170-194. Carroll. M. L. & D. S. Wethey. 1990. Predator foraging behaviour: effect of a novel prey species on prey selection by a marine intertidal gas- tropod. J. E.xp. Mar. Biol. Ecol. 139:101-117. Debrot. A. 1990. Temporal aspects of population dynamics and dispersal behaviour of the West Indian topshell Ciirarium pica, at selected sites in the Exuma Cays. Bahamas. Bull. Mar. Sci. 47(2):431^t47. Ellison. A. M. & E. J, Famswonh, 1992, The ecology of Belizean man- grove root fouling communities: patterns of epibiont distribution and abundance, and effects on root growth. Hydrobiologia 247(l-3):87- 98. Fairweather. P. G. 1988. Movements of mtertidal whelks {Morula mar- ginalba and Thais orbita) in relation to availability of prey and shelter. Mar. Biol. 100:63-68. Fairweather. P. G.. A. J, Underwood & M. J. Moran. 1984. Preliminary investigations of predation by the whelk Morula marginalba. Mar. Ecol. Prog. Ser. 17:143-156. Garrity. S. D, 1984. Some adaptations of gastropods to physical stress on a tropical rocky shore. Ecology 65(2):559-574. Gayanilo, F. C. P. Span-e & D. Pauly. 1994. The FAO-ICLARM Stock Assessment Tools (FiSAT) user's guide. Final draft. FAO. Rome. ICLARM Contr. 1048. Geller. J. B. 1990. Consequences of a morphological defense: growth. repair and reproduction by thin- and thick-shelled morphs of Nucella cmarginata (Deshay es). J. E.xp. Mar. Biol. Ecol. 144:173-184. Gocke. K.. M. Vitola & G. Rojas. 1981. Oxygen consumption patterns in a mangrove swamp on the Pacific Coast of Costa Rica. Rex. Biol. Trap. 29:143-154. Gulland. J, A, 1969, Manual of methods lor fish stock assessment. Part 1, Fish population analysis, FAO Man. Fish. Sci. 4. 154 pp. Janzen. D. H. (ed.). 1991, Hislona natural de Costa Rica, ISBN 9977- 67-169-9. Editonal de la Umversidad de Costa Rica. San Jose. Costa Rica. 822 pp. Jerome. T. G. R. L. 1987. Vertical distributions, shell morphologies and predation of mangrove snails. Associated Colleges of the Midwest. 34 pp. Keen. A. M. 1971. Sea Shells of Tropical West America. Manne Mol- lusks from Baja California to Peru. 2nd ed, Stanford University Press, Stanford. California. 1064 pp. Kitching, J. A. 1977. Shell form and niche occupation in Nucella lapillus. J. Exp. Mar. Biol. Ecol. 26:275-289. Kitching, J. A. 1986. The ecological significance and control of shell variability in dogwhelks from temperature rocky shores, pp. 234-248. In: P. G. Moore and R. Seed (eds). The Ecology of Rocky Coasts. Columbia University Press. New York. Lalana Rueda. R. & F. Gosselck. 1986. Investigations of the benthos of mangrove coastal lagoons in southern Cuba, Int. Rev. Gesamt. Hydro- biol. 71(6):779-794. Menge. B. A. 1978. Predation intensity in a rocky intertidal community. Effect of algal canopy, wave action and desiccation on predator feeding rates. Oecologia 34:17-35. Moore. H. B. 1972. Aspects of stress in the tropical marine environment. Adv. Mar. Biol. 10:217-269. 432 Koch and Wolff Munro. J. L. 1982. Estimation of the parameters of the von Bertalanffy growth equation from recapture data at variable time intervals. J . Cons. CIEM. 40:199-200. Palmer. R. A. 1979. Fish predation and the evolution of gastropod shell sculpture: experimental and geographic evidence. Evoliilion 33(2): 697-713. Palmer, R. A. 1981 . Do carbonate skeletons limit the rate of body growth? Nature 292A50-\52. Palmer. R. A. 1984. Prey selection by thaidid gastropods: some observa- tional and experimental field tests of foraging models. Oecologia 62: 62-72. Palmer, R. A. 1985. Adaptive value of shell variation in Thais lameUosa: effect of thick shells on vulnerability to and preference by crabs. Vetiger 27(4):349-356. Palmer, R. A. 1992. Calcification in marine molluscs: how costly is it? Proc. Nail. Acad. Sci. USA 89(4):1379-I382. Pauly, D. 1984. Fish Population Dynamics in Tropical Waters: A Manual for Use With Programmable Calculators. Studies and Reviews. ICLARM. Philippines. 325 pp. Pauly, D. & G. Gaschiitz. 1979. A simple method for fitting oscillating length growth data, with a program for pocket calculators. I.C.E.S. CM. 1979/G. Demersal Fish Cttee. 24. 24 pp. Pauly, D. & J. L. Munro. 1984. Once more on the companson of growth in fish and invertebrates. Fishbyte 2:21. Perry. D. M. 1988. Effects of associated fauna on growth and productivity in the red mangrove. Ecology 69(4):1064-1075. Peterson, C. L. I960. The physical oceanography of the Gulf of Nicoya, a tropical estuary. Bull. InleiAmerican Trap. Tuna Commission 4(4), 37 pp. Prince, J. D., T. L. Sellers, W. B. Ford & S. R. Talbot. 1988. Recruit- ment, growth, mortality and population structure in a southern Aus- tralian population of Haliolis rubra iMollusca: gastropoda). Mar. Biol. 100:75-82. Richardson. T. D. & K. M. Brown. 1990. Wave exposure and prey size selection in an intertidal predator. J. Exp. Mar. Biol. Ecol. 142(1-2): 105-120. Russell-Hunter. W. D. 1985. Physiological, ecological and evolutionary aspects of molluscan tissue degrowth. Am. Malacol. Bull. 3(2):2L3- 221. Sainsbury. K. J. 1982. Population dynamics and fishery management of the paua, Haliotis ins. I. Population structure, growth, reproduction and mortality. N.Z. J. Mar. Freshwater Res. 16:147-161. Spight. T. M. 1983. Risk, reward and the duration of feeding excursions by a marine snail. Veliger 24(4):302-308. Thomas. P. 1988. Die Niederschlagsverhaltnisse in Costa Rica. Master's Thesis. University of Augsburg. Augsberg, Germany, 88 pp. Underwood. A. J. 1978. The ecology of intertidal gastropods. Adv. Mar. Biol 16:111-210 Vakily, J. M. 1992. Determination and comparison of growth in bivalves, with emphasis on the tropics and Thailand. ICLARM Tech. Rep. 36: 125. Vermeij, G. J. 1972. Intraspecific shore-level size gradients in molluscs. Ecology 53:693-700. Vermeij, G. J. 1978. Biogeography and adaptation. Harvard University Press, Cambridge, Massachusetts. 332 pp. Voorhis. A. D., C. E. Epifanio, D. Maurer, A. I. Dittel & J. A. Vargas. 1983. The estuarine character of the Gulf of Nicoya, an embayment on the Pacific Coast of Central America. Hydrobiologia 99:225-237. Wellington, G. M. & A. M. Kuris. 1983. Growth and shell variation in the tropical Eastern Pacific intertidal gastropod genus Purpura: eco- logical and evolutionary implications. Biol. Bull. 164:518-535. West. L. 1988. Prey selection by the tropical snail Thais melones: a study of interindividual variation. Ecology 69(61:1839-1854. Withey. J. 1990 Predation and Vertical Distribution of Four Snail Species in a Tropical Mangrove Estuary. Associated Colleges of the Midwest, 38 pp. Wolff, M. 1985. Fischerei. Okologie und Populationsdynamik der Pilger- muschel Argopeclen purpuratus (L.) im Fischereigebiet von Pisco (Peru) unter dem Ein-flup des El Nino 1982/83. Ph.D. Thesis, Uni- versity of Kiel, 113 pp. Wolff, M. 1987, Population dynamics of the peruvian scallop Argopecten purpuratus during the El Nino Phenomenon of 1983. Can. J. Fish. Acjuat. Sci. 44:1684-1691. Wolff, M. 1989. Estimates of growth, mortality and recruitment of the Loco Con cho-lepas concholepas (Bruguiere, 1789) derived from a shell mound in northern Chile. Stud. Neotrop. Fauna Em-. 24(2):87- 96. Wolff, M. 1994. Population Dynamics, Life Histories and Management of Selected Invertebrates of the SE Pacific Upwelling System. Habilita- tion at the University of Bremen. Bremen, Germany. 210 pp. Wu, Q.. Z Fengwu. L. Lin, J. Jingxiang, L. Rongguan, C. Erxi, H. Minghai & H. Xinguang. 1992. Benthic communities of mangrove area in Daya Bay. J. Oceanogr. Taiwan Strait Taiwan Haixia 11(2): 161-166. Journal of Shellfish Resenrch. Vol, 15. No. 2. 4.V^^35. 1996. EVALUATION OF THREE ANESTHETIC AGENTS FOR CRAYFISH (ORCONECTES VIRILIS) P. B. BROWN,' * M. R. WHITE," J. CHAILLE,' M. RUSSELL,' AND C. OSETO' ^Depariment of Forestry and Natural Resources '^Animal Disease Diagnostic Laboratory, and ^Department of Entomology Purdue University West Lafayette. Indiana 47907 ABSTRACT Crayfish are important research animals, and their culture m Louisiana and neighbonng states is one of the largest aquacultural mdustries in the United States; however, there are no proven anesthetic agents for use in crayfish. In this study, we compared tricaine methane sulfonate (MS-2221. lidocaine-HCl. and ketamine-HCl as anesthetic agents for Orconectes virilis. MS-222 was ineffective for crayfish at dosages as high as 1,000 mg/L, applied as a bath treatment. Lidocaine-HCl and ketamine-HCl were administered either intramuscularly (IM) or intrathoracically (IT). IT injections of either agent resulted in inconsistent anesthetization. IM injections of both agents resulted m anesthetization. The minimum effective dose of lidocainc-HCl was greater than M\0 |jLg/g body weight. The duration of anesthetization at the higher doses was between 20 and 25 min. Ketamine was effective at doses of more than 90 |i.g/g body weight with durations of longer than 1 h. No mortalities occurred during any of the evaluations. On the basis of these data, both lidocaine- and ketamine-HCl are efficacious in crayfish. Lidocaine-HCl can be used for short-term anesthetization, whereas the effects of ketamine-HCl were long term. KEY WORDS: Cravfish, anesthetic, MS-222, lidocaine, ketamine INTRODUCTION Tricaine methane sulfonate (MS-222). applied as a bath treat- ment, is the only anesthetic agent approved for use in aquatic animals in the United States. It is commonly used to anesthetize fish for research purposes and is routinely used in practical settings to minimize physiological responses to handling. However, the evaluation of anesthetics for use in freshwater crustaceans is rare, although a significant commercial industrv' exists and freshwater crayfish arc routinely used as research animals. One of the most common methods of immobilizing crustaceans is to place individuals in cold water or freezers before the collec- tion of pertinent measurements or tissues. Recommendations for the anesthetization of marine decapods include chloroform, ethane disulphonate, fresh water, heat (40°Cl, isobutyl alcohol, methyl pentynol, procaine-HCl, and soda water (Smaldon and Lee 1979). Most of these methods result in slow and inconsistent immobili- zation, particularly for those applications that may require the rapid acquisition of tissues for physiological measurements In this study, we evaluated three separate anesthetic agents for their use- fulness in freshwater crayfish. The anesthetics evaluated were tri- caine methane sulfonate, lidocaine-HCl, and ketamine-HCl, MATERIALS AND METHODS Tricaine methane sulfonate was added to individual closed con- tainers so that concentrations were 0, 50, 100, 500, or 1,000 mg/L. Two adult male Orcoiiectes virilis were placed in each container. All crayfish were intermolt and Form I. Individual weights of crayfish ranged from 33 to 37 g. Crayfish were accli- mated to normal laboratory temperatures {22-26°C). and all eval- uations were conducted within that range of temperatures. Cray- fish were monitored for 30 min after the onset of anesthetic de- livery to evaluate effects. All crayfish were given tactile Corresponding author; Dr. Paul Brown, Purdue University, 1 159-Forestry Buildms, West Lafayette, IN 47907-1159. Stimulation approximately every 30 s during this period. If indi- vidual crayfish were not anesthetized by 5 min, the dosage was considered insufficient. This entire study was conducted twice. Lidocaine-HCl and ketamine-HCl were injected into male cray- fish of varying weights (27—1-5 g). Injections were either intra- muscularly (IM) into the tail or intrathoracically (IT). IM injec- tions were into the third sternum posterior to the thorax, and IT injections were into the base of the fifth periopod and into the thorax proper. Once the needle was inserted, hemolymph was aspirated for verification of location. Lidocaine-HCl (20 mg/mL) was injected in the following vol- umes; O.OI , 0,025. 0.05, 0. 10, 0.20, and 0.50 mL. Ketamine-HCl was acquired as a concentration of 100 mg/mL. In the first at- tempt, the stock solution was used, the administration was IM and the injection volume was 0.025 mL. In further evaluations, the stock solution was diluted to 20 mg/mL with distilled water im- mediately before the study was started. The diluted solution was then injected in the following volumes: 0.01, 0.025, 0.05, and 0.10 mL. After injection by either route, individual crayfish were monitored as described above. The time to unresponsiveness to tactile stimuli was recorded as the time to anesthetization; then, duration of anesthetization was monitored. Before the plotting of data, all doses were expressed as micrograms of anesthetic per gram of crayfish. All crayfish exposed to MS-222 and those injected by either route with lidocaine-HCl were placed in a recovery tank. All cray- fish administered ketamine-HCl were euthanized by hypothermia after complete recovery. RESULTS AND DISCUSSION Tricaine methane sulfonate, regardless of concentration, had no anesthetic effect in either study. Of the crayfish injected IT with lidocaine-HCl, only 6 of 24 became unresponsive to tactile stim- uli. Only one test animal (25%) became unresponsive when given volumes of 0,05, 0,2, and 0.5 mL, whereas 75% of those injected with 0.10 mL were anesthetized. The average time to unrespon- 433 434 Brown et al. 2:09:36 -r 1:55:12-- u 1:40:48-- 1:26:24 - ■ 1:12:00- ^ 0:57:36 ■ - c 0:43:12 • ■ ^ 0:28:48 ■ • 0:14:24 - - 0:00:00 ug Ketamine/g crayfish Figure 1. Mean duration of anesthetization of crayPish recei\in)i IM Injections of lidocaine-HCI. siveness was 30 s, and the average duration was 5 min 48 s; the maximum duration of IT lidocaine-HCI anesthetization was 8 min 42 s. IM injections of lidocaine-HCI had no effect on crayfish until the injection volume was 0. 10 mL. Fifty percent of those injected with 0.10 and 0.20 mL became unresponsive, whereas all of those injected with 0.50 mL were anesthetized. Average time to anes- thetization was 1.5 min. with a ma.\imum duration of longer than 25 min. The response of crayfish to IT injections of ketamine was highly variable. Only 5 of 16 crayfish became unresponsive. The duration of anesthetization with IT injection was also highly vari- 25:55 T 23:02 ■• 20:10-- 1 17:17 c o 14?4 N V C V 11:31 c 10 ^*- o c 08:38 5 3 Q 05:46 02:53 ■ • 00:00 ■+■ -f- 16 20 ■+■ -+■ S 31 43 64 ug Lidocajne/g crayfish Figure 2. Mean duration of anesthetization of crayfish receiving IM injections of ketamine-HCl. 500 Anesthetic Agents for Crayfish 435 able. Two of the crayfish remained unresponsive for over 2 h but became responsive and appeared normal after that time, whereas two individuals receiving the same dose did not exhibit any effects of the anesthetic. Crayfish receiving ketamine IM exhibited the most consistent anesthetic response. All crayfish receiving injection volumes of 0.05 mL and higher of the diluted solution (20 mg/mL) and all of those receiving 0.025 mL of the stock solution were anesthetized. The overall average time to anesthetization was 54 s. The duration of unresponsiveness increased from an average of 8 min 48 s in crayfish injected with 0.05 mL of the diluted solution to 1 h 48 min crayfish injected with 0.025 niL of the stock solution. The response of crayfish administered IT injections of both lidocaine and ketamine was inconsistent and highly variable. The reasons for this remain unclear. Thus, that route of administration cannot be recommended. The responses of crayfish administered IM injections of both anesthetics were more consistent. On the basis of the graphical presentation of the duration of anesthetization as a function of dose (micrograms per gram cray- fish body weight), the minimum effective dose of lidocaine-HCl for the consistent immobilization of O. virills is more than 300 jjLg/g (Fig. 1). We recommend the IM injection of doses of 400 |j.g/g for the light anesthetization of crayfish. Lidocaine-HCl has been used as an anesthetic in fish, but as a bath treatment similar to the administration of MS-222 (Carrasco et al. 1984). It seems clear from Figure 2 that ketamine-HCl is a more potent anesthetic than lidocaine-HCl. Once doses exceed approximately 40 (JLg/g, consistent anesthetization of longer than 10 min oc- curred. Crayfish receiving doses in excess of 85 jjig/g were in a plane of deep anesthesia but recovered completely. Thus, dosages of ketamine-HCl of more than 90 iJLg/kg IM are recommended for deep anesthetization. Ketamine-HCl is approved for felines and subhuman primates and is used in several other species (Alexander 1985, Booth and McDonald 1988). Ketamine-HCl can cause ad- verse emergence reactions such as hallucinations, confusion, and irrational behavior (Brown 1993). When fish were administered ketamine-HCl at doses above 45 mg/kg, emergence reactions were observed (Williams et al. 1988). Crayfish administered ketamine- HCl in these studies were fully recovered before euthanasia. Re- covery was similar to those administered lidocaine-HCl in that no adverse reactions were noted. ACKNOWLEDGMENTS This project was partially supported by the National Science Foundation, Young Scholars Program. The project was also sup- ported by the Purdue Agricultural Research Programs (IND 059054). We thank the staff at the Aquaculture Research Facility for their help. This article is technical contribution number 1492 1 , Purdue University Agricultural Research Programs. LITERATURE CITED Alexander, F. 1985. An Introduction to Veterinary Pharmacology. 4th cd. Churchill Livingstone, Inc., New York- Booth, N. H. & L. E. McDonald. 1988. Vetennary Pharmacology and Therapeutics. 6th ed. Iowa State University Press. Ames. Iowa. Brown. L. A. 1993. Anesthesia and restraint, pp 79-90. In: M. K. Stoskopf (ed.). Fish Medicine. W.B. Saunders Co., Philadelphia. Carrasco, S.. H. Sumano & R. Navarro-Fierro. 1984. The use of lidocaine-sodium bicarbonate as anesthetic in fish. Aijiuwullure 41: 395-398. Smaldon. G. & E. W. Lee. 1979, A Synopsis of Methods for the Nar- cotisation of Marine Invertebrates. Royal Scottish Museum Informa- tion Series. Natural History 6, Edinburgh, Scotland. 96 pp. Williams, T. D . J. Chnstiansen & S. Nygren. 1988. A comparison of intramuscular anesthetics in teleosts and elasmobranchs. /«/. Assoc. Aquat. Aiiim. Med. Proc. 19; 148. Journal of Shellfish Research. Vol. 15, No. 2. 437-440. 1996. ELECTROPHORETIC DATA SUPPORT THE LAST-MALE SPERM PRECEDENCE HYPOTHESIS IN THE SNOW CRAB, CHIONOECETES OPILIO (BRACHYURA: MAJIDAE) JEAN-MARIE SEVIGNY AND BERNARD SAINTE-MARIE Direction clcs inveriebres el de la hiologie e.xperimentale Institiit Maurice-Lamontagne Ministere des Peches el des Oceans 850 route de la Mer C.P. WOO. Mont-Joli (Quebec) G5H 3Z4, Canada ABSTRACT Controlled mating experiments were carried out with the snow crab. Chionoecetes opilio. over two female breeding cycles. In this species, electrophoretic patterns of the enzymes glucose-6-phosphate isomerase. phosphoglucomutase. and malate dehydrogenase observed in the parents and the progenies are inherited as simple Mendelian codominant characters. The genetic differences observed between the larvae from the first and second egg clutches of some females mated with males of different genotypes support the hypothesis that the last male to mate w ith a female before she extrudes an egg clutch fertilizes a large proportion of the extruded eggs. KEY WORDS: Snow crab. Chionoecetes. allozyme. electrophoresis, sperm competition, mating experiments INTRODUCTION Few studies have addressed the question of sperm competition in Crustacea. Polyandry and sperm mixing leading to multiple paternity seem to be the rule in two terrestrial isopods. Porccllio scaber (Sassaman 1978) and Venezillo everghuleiisis (Johnson 1982). Multiple paternity is relatively rare in wild populations of the amphipod Gommarus (Sexton 1935, Yamold 1935a, Yamold 1935b, Siegismund 1985), where precedence of the first male mate has been attributed to the fact that sperm is not stored be- tween broods and that fertilization occurs <3 h after copulation (Birkhead and Pringle 1986). However, there appears to be no general trend regarding sperm competition among Decapoda. Al- though electrophoretic data revealed sperm mixing and multiple paternity in the lobster Homonis americunus ( Nelson and Hedge- cock 1977) and in the crab Cancer pagunis (Burfitt 1980), serial mating experiments carried out with irradiated and normal males showed that the last male mate fertilizes a very large proportion of extruded eggs in the crab Scopimera globo.sa (Koga et al. 1993) and in the crayfish Orconectes ruslicus (Snedden 1990). Last-male sperm precedence has also been observed in the majid Inachus phalangiurn (Diesel 1990). In the snow crab. Chionoecetes opilio. there is a potential for sperm competition to occur within the spemiatheca at any stage of the female's reproductive life. Indeed, sperm competition might occur at the primary spawning, because females can copulate with more than one male before extruding their first clutch of eggs (Sainte-Maric et al. unpub. obs.). During their first mating period, females generally receive more than enough spenn to fertilize their first egg clutch, and excess sperm may remain viable in long-term storage within the spermathecae (Sainte-Marie and Lovrich 1994, Sainte-Marie and Carriere 1995). Because females may reniate before extruding their second or ulterior egg clutch (Taylor et al. 1985, Conan and Comeau 1986. Hooper 1986), there also exists the possibility of competition between older stored sperm and that acquired more recently in the reproductive life of the female (Elner and Beninger 1992, Elner and Beninger 1995). Beninger et al. ( 1991 ) and Elner and Beninger ( 1992) proposed for C . opilio that the most recently acquired sperm has precedence over older sperm in fertilizing the next egg clutch. However, last- male sperm precedence has yet to be demonstrated in this species by the use of controlled mating experiments (Elner and Beninger 1995). In this article, we provide for C. opilio genetic data sup- porting the hypothesis of last-male sperm precedence based on mating experiments carried out over two female breeding cycles. MATERIALS AND METHODS Collection and Mating of Crabs Males and immature females used in mating experiments were collected in Bale Sainte-Marguerite (ca. 50°06'N, 66°33'W), northwest Gulf of Saint Lawrence, in September-October 1991. All crabs were identified individually with a numbered plastic tag attached to the basipodite of the fourth or fifth pereiopod. Adult males and immature females were kept in separate holding tanks, and the females were checked daily for molting. Less than 12 h after an immature female molted to maturity, which occurred from 8 January to 12 April, 1992, she was introduced to one hard- shelled adult male in a I20-L tank. After mating and hardening, the female was transferred to a female communal holding tank and maintained on a diet of previously frozen shrimp and herring. For those females that extruded fertilized eggs after mating, embryonic development lasted approximately 1 y at a mean temperature of l.5°C. Berried females were checked periodically, and before the eggs hatched, they were transferred to individual tanks. Hatching periods for individual females lasted from 2 to 24 d. All larvae were collected every 24 h. Several larvae were placed on glass fiber filters on which the hatching date was noted, rinsed in dis- tilled water, and immediately frozen in liquid nitrogen. They were stored at - 80°C for subsequent electrophoretic analysis. Twenty of 54 berried females were given the opportunity to remate at the hatching of their first brood, as one male different from their initial mate was introduced into their hatching tank. The females presumably remated (Sainte-Marie and Carriere 1995) and, after hatching their first egg clutch and extruding a second one, were returned to the female communal tank. For various reasons, only 13 of the remated females were used in this genetic study. Embryos were sampled from different places in the second egg clutch after 3 and 8 mo of incubation. These embryos were 437 438 SfiVIGNY AND SaINTE-MaRIE handled and preserved as described above for the larvae. Samples of muscle tissue were collected from the second pereiopod of all male and female parents and stored at - 80°C until electrophoretic analysis was carried out. Spermathecae of the females given the opportunity to remate were dissected to determine if they con- tained fresh ejaculate, confirming that the females had effectively remated. Recent ejaculate appeared as a pale deposit underlying the darker and + 1-y-old ejaculates. Electrophoretic Analysis Approximately 100 mg of leg muscle tissue from each of the parents was homogenized in a 0.01 M Tris-HCL (pH 8.0) extrac- tion solution containing 30% sucrose, 0.005 M dithiothreitol, and 0.5% polyvinylpolypyrolidone. Homogenates were centrifuged at 15,000 g for 60 min at 4°C. First-clutch larvae hatched on different days and second-clutch embryos samples after 3 and 8 mo of incubation were homogenized individually with a pipette tip di- rectly in sample well plates containing 5 to 10 |jiL of the same extraction buffer. The electrophoretic patterns of the enzymes glu- cose-6-phosphate isomerase (GPI, EC 5.3.1 .9), phosphoglucomu- tase (PGM, EC 5.4.2.2). and malate dehydrogenase (MDH, EC 1 . 1 . 1 .37) were studied by use of the cellulose acetate gel electro- phoresis techniques of Hebert and Beaton (1989). RESULTS Allozyme variation is low in C. opilio (Davidson et al. 1985, Sevigny unpub. obs.), and in our experiments, two alleles segre- gated at the MDH* and PGM* loci, whereas three were detected at the GPI* locus. For any given clutch of eggs, there was no difference in the electrophoretic patterns of larvae with respect to hatching date, so data are pooled in subsequent analyses. In 17 of the 62 tested crosses, the male and the female geno- types differed at least at one of the three studied loci, and in two crosses (#3 and #11), they differed at both the MDH* and the PGM* loci (Table 1). For all crosses, the observed number of genotypes of the progeny was in good agreement with that ex- TABLE 1. Inheritance of the electrophoretic patterns for the enzymes MDH, GPI, and PGM in the progenies of monandrous female C. opilio. Genot) pes in Paired Mating Female Male Larvae No. MDH GPI PGM MDH GPI PGM MDH GPI PGM 1 alal alal alal alal alal ala2 21 alal 21 alal II (10.5) alal I0(10.5)ala2 2 alal alall ala2 alal alal alal 21 alal 20 alal 21 (21.5) alal 22(21.5)ala2 3 ala2 alal alal alal alal ala2 12(12.0) alal 12(12.0) ala2 22 alal 10 (10,5) alal 11 (10.5) ala2 4 alal alal alal alal alal ala2 21 alal 10 alal 10 (10.0) alal 10(10.0) ala2 5 alal alal ala2 alal alal alal 10 alal 10 alal 12 (11.0) alal 10(11.0) ala2 6 alal alal alal alal ala3 alal 11 alal 6 (5.5) alal 5 (5.5) ala3 1 1 alal 7 alal alal alal alal alal ala2 11 alal 11 alal 12(11 0) alal 10(1 1.0) ala2 8 alal alal alal alal alal ala2 11 alal 11 alal 11 (11.0) alal 11 (11.0) ala2 9 alal alal alal alal alal ala2 10 alal 10 alal 109 (106.5) alal 104 (106.5) ala2 10 alal alal alal alal alal ala2 11 alal 11 alal 23 (22.0) alal 21 (22.0) ala2 11 ala2 alal alal alal alal ala2 10(11.0) alal 12(1 1.0) ala2 1 1 ala2 10 (11.0) alal 12(11.0)ala2 12 alal alal alal alal alal ala2 10 alal 9 alal 68 (66.0) alal 64 (66.0) ala2 13 alal alal alal alal ala3 alal 10 alal 10(10.5) alal 11 (10.5) ala3 10 alal 14 alal alal alal alal ala3 alal 10 alal 21 (21.5) alal 22(21.5)ala3 21 alal 15 alal alal ala2 alal alal alal 10 alal 11 alal 10 (10.5 alal 11 (10.5) ala2 16 alal alal alal ala2 alal alal 6 (6.0) alal 6 (6.0) ala2 11 alal 1 1 alal 17 alal alal alal alal ala2 alal 11 alal 12 (11.0) alal 22 alal 10(11.0) ala2 The numbers for each genotype expected for Mendelian codominant characters are shown in parentheses for the larvae of each cross. Last-Male Sperm Precedence in C. opilio 439 TABLE 2. Inheritance of the electrophoretic patterns for the enzymes MDH, GPI, and PGM in two successive progenies of biandrous female C. opilio with mates of different genotypes. Genotypes of First Mates Genotypes of Second Mates Female Male Cross — No. MDH GPI PGM MDH GPI Larvae Male Larvae PGM MDH GPI PGM MDH GPI PGM GPI PGM NT 78 alal 0ala2 NT 73 alal OalaZ 71 alal NT 0 ala3 M alal alal alal 12 alal alal alal 14 alal alal alal alal ala3 111 alal lU alal 10 alal 4 alal HI alal 21 (21.5)alal 22(21.5)ala3 109 (106,5) alal alal alal alal 104 (106 .5) ala2 68 (66 0) alal alal alal alal 64 (66-0) ala2 21 alal alal alal alal The numbers of each genotype expected for Mendelian codominant characters are shown in parentheses tor the larvae of each cross. Family numbers refer to those in Table 1 . NT. not tested. pected for Meniiclian codominant characters (p > 0.67 in the 19 X" tests of goodness of fit). The offspring of a homozygous indi- vidual mated with a heterozygous one are expected to segregate in a 1 : 1 ratio, as was observed in these mating experiments (Table I ). Among the 13 females that were presented with a second mate before extrusion of their second egg clutch. 1 1 had recent ejaculate in their sperinathecae. Indicating that they had remated with the second male. Of these 1 1 females. 3 had second mates whose genotype differed from that of the first male parent (Table 2). The three females were homozygous at the three tested loci, whereas their first male mates were heterozygous at the PGM* (crosses #9 and #12) and GPI* (cross #14) loci. Their second mates were all homozygous at the three loci and were of the same genotypes as the females. The genotypes of the larvae in the first and second clutches of each fetnale differed. Indeed, the heterozygotes that were observed at the PGM* locus in crosses #9 and #12 and at the GPI* locus in cross # 14 (Tables I and 2) were not detected in larvae from the second clutch, which were all homozygous at the three studied loci (Table 2). Such a result would be expected if the second male mate had sired most or all of the eggs in the second clutch. DISCUSSION The first mating experiment shows that MDH. GPI. and PGM electrophoretic patterns are inherited as Mendelian codominant characters and that ontogenetic changes are not important factors in our experiments. The absence of heterozygous individuals in the second egg clutch of all three females of the second mating ex- periment indicates that, under the laboratory conditions presented here, the last male to mate with a female before she extrudes eggs sires most or all of the progeny. Although a contribution of the sperm from the first male mate to the pool of homozygotes from the second clutch would not be detected, the absence of heterozy- gotes in the second clutch indicates that such a contribution, if any, is very limited. The fact that 32 of 34 females that were not remated extruded a fertilized second egg clutch (Sainte-Marie and Carriere 1995) strongly suggests that females that were remated still had viable stored sperm remaining over from their first mat- ing. Our results are in agreement with the hypothesis that sperm from the last male has precedence in fertilizing eggs in C. opilio (Beninger et al. 1991, EIner and Beninger 1992, Elner and Be- ninger 1995). However, our experimental design does not allow us to make any inference regarding the mechanisms that might lead to last-male sperm precedence in C. opilio. The observation that the first gonopod of male C. opilio is hook shaped led to the hypoth- esis that males can extract previously stored ejaculate from the spermathecae (Beninger et al. 1991, Elner and Beninger 1992, Elner and Beninger 1995). This mechanism differs from that de- scribed in another majid, /. phalangium, where last-male sperm precedence is ensured by the displacement of previous sperm de- posits deeper into the spermatheca (Diesel 1988, Diesel 1990). Sperm displacement resulting in last-male sperm precedence has also been predicted or demonstrated to occur in other brachyurans with ventral-type spermathecae (Murai et al. 1987, Diesel 1988. Diesel 1990, Koga et al. 1993. Orensanz et al. 1995). including Cluonoecetes hairdi (Paul et al. 1983). Further laboratory exper- iments will be necessary to elucidate the mechanisms underlying last-male sperm precedence in C. opilio. Our study does not rule out the possibility that multiple pater- nity can occur under some circumstances for C . opilio. For in- stance, a female that extrudes a second clutch of eggs without remating might draw on sperm reserves obtained from different males during the first mating cycle. Another case that would most certainly lead to multiple paternity would anse when a female is taken over by another male during oviposition and is then insem- inated anew, as may happen (Sainte-Marie unpub. obs.). Further investigations are currently under way to describe the ecological contexts that might favor single or multiple paternity. ACKNOWLEDGMENTS We thank E. Parent. F. Hazel. Y. Gauthier. P. Goudreau. J. Quimper. M. Belanger. and M. Champagne for their assistance in the field and the laboratory. This work was funded by the Quebec Federal Fisheries Development Program. LITERATURE CITED Beninger. P. G., R. W. Elner. & Y. Poussart. 1991. The gonopods of the majid crab Cluonoecetes opilio (0. Fabricius). J. Crustacean Biol. lI;217-228. Birkhead. T. R. & S Pringle. 1986. Multiple mating and paternity in Gammarus pulex. Anim. Behav. 34:611-613. Burfitt. A. H. 1980. Glucose phosphate isomerase in Cancer pagurus L. 440 StVIGNY AND SaINTE-MaRIE broods as evidence of multiple paternity (Decapoda. Brachyura). Crus- laceana 39-306-3>\0. Conan, G. Y. & M. Comeau. 1986. Functional maturity and terminal molt of male snow crab. Chionoecetes opilio. Can. J. Fish. Aqiiat. Sci. 43:171(3-1719. Davidson. K., J. C. Roff & R. W. Elner. 1985. Morphological, electro- phoretic, and fecundity characteristics of Atlantic snow crab. Chiono- ecetes opilio. and implications for fisheries management. Can. J. Fish. Aquat. Sci. 42:476-482. Diesel. R. 1988. Discrete storage of multiple-mating sperm in the spider crab Inachus phalangiwn. Nalurwissenschaflen 75:148-149. Diesel. R. 1990. Sperm competition and reproductive success in the de- capod Inachus phalangium (Majidae): a male ghost spider crab that seals off rivals" sperm. J. Zool. Land. 220:213-223. Elner, R. W. & P. G. Beninger. 1992. The reproductive biology of snow crab, Chionoecetes opilio: a synthesis of recent contributions. Am. Zool. 32:524-533. Elner. R. W. & P. G. Beninger. 1995. Multiple reproductive strategies in snow crab. Chionoecetes opilio: physiological pathways and behavior- al plasticity. J. Exp. Mar. Biol. Ecol. 193:93-112. Hebert. P. & M. Beaton. 1989. Methodologies for Allozyme Analysis Using Cellulose Acetate. Practical Handbook. Educational Service of Helena Laboratories. Beaumont. Te.xas. Hooper, R. G. 1986. A spring breeding migration of the snow crab. Chio- noecetes opilio (O. Fabricius). into .shallow water in Newfoundland. Crustaceana 50:257-264. Johnson, C. 1982. Multiple insemination and sperm storage in the isopod, Venezillo evergladensis Schultz. 1963. Crustaceana 42:225-232. Koga. T.. Y. Henmi & M. Murai. 1993. Sperm competition and the assurance of underground copulation in the sand-bubbler crab Scopi- mera globosa (Brachyura: Ocypodidae). J . Crustacean Biol. 13:134— 137. Murai. M.. S. Goshima & Y. Henmi. 1987. Analysis of the mating system of the fiddler crab. Uca lactea. Anim. Behav. 35:1334-1342. Nelson, K. & D. Hedgecock. 1977. Electrophoretic evidence of multiple paternity in the lobster Homarus americanus (Milne-Edwards). Am. Nat. 11:361-365. Orensanz. J. M.. A. M. Parma. D. A. Armstrong. J. Armstrong & P. Wardrup. 1995. The breeding ecology of Cancer .gracilis (Crustacea: Decapoda: Cancridae) and the mating systems of cancrid crabs. J. Zool. 235:411-437. Paul, A. J., A. E. Adams, J. M. Paul & H. H. Feder. 1983. Some aspects of the reproductive biology of the crab Chionoecetes bairdi. University of Alaska Sea Grant Rep. 83-1. Sainte-Marie. B. & C. Carriere. 1995. Fertilization of the second clutch of eggs of snow crab. Chionoecetes opilio. mated once or twice after their molt to maturity. Fish. Bull. U.S. 93:758-763. Sainte-Marie. B. & G. A. Lovrich. 1994. Delivery and storage of sperm at tlrst mating of female Chionoecetes opilio (Brachyura. Majidae) in relation to size and morphometric maturity of male parent. J. Crusta- cean Biol. 14:508-521. Sassaman. C. 1978. Mating system in porcellionid isopods: Multiple pa- ternity and sperm mixing in Porcellio scaher Latr. Heredity 41:385- 397. Sexton. E. W. 1935. Fertilization of successive broods of Gammarus chevreu-xi. Nature 136:477. Siegismund. H R. 1985. Genetic studies oiGammarus. Ill Inheritance of electrophoretic variants of the enzyme mannose phosphate isomerase and glucose phosphate isomerase in Gammarus oceanicus. Hereditas 102:25-31. Snedden. W. A. 1990. Determinants of male mating success in the tem- perate crayfish Orconecles ruslicus: chela size and sperm competition. Behaviour 115:100-113. Taylor. D. M.. R. G. Hooper & G. P Ennis. 1985. Biological aspects of the spring breeding migration of snow crabs. Chionoecetes opilio, in Bonne Bay. Newfoundland (Canada). Fish. Bull. U.S. 83:707-711. Yamold. K. W. 1935a. Further reappearance of the second red-eye mu- tation in Gammarus. Nature 135:832-833. Yamold, K. W. 1935b. Persistence of sperms to a later mating in Gam- marus. Nature 136:758-759. Journal of Shellfish Research. Vol. 15. No. 2, 441-445, 19%. COMPARISON OF EXCHANGE AND NO-EXCHANGE WATER MANAGEMENT STRATEGIES FOR THE INTENSIVE POND CULTURE OF MARINE SHRIMP J. STEPHEN HOPKINS,* PAUL A. SANDIFER, CRAIG L. BROWDY, AND JOHN D. HOLLOWAY Waddell Mahculturc Center P.O. Box 809 Bluff ton. South Carolina 29910 ABSTRACT Most of the potential and realized adverse environmental effects of shrimp farming are associated with routine water exchange. This study compared shrmip production and water quality in tnplicate ponds operated with and without water exchange. No statistical differences were detected in growth or survival among treatments, although there was a trend towards slightly smaller mean size at harvest and lower survival in the ponds operated without water exchange. The ponds operated with and without routine water exchange had average production of 5,888 and 5.444 kg/ha per crop, respectively. Differences in harvest size and survival also influenced food conversion efficiency. The ponds operated without water exchange had higher nutrients and biochemical oxygen demand (BOD) at the end of the study and. thus, discharged more nutrients and BOD in the drain harvesting process. However, the continuous discharge from the ponds operated with water exchange probably resulted in a much larger total nutrient and BOD load to the adjacent estuary. Heavy precipitation resulted in higher turbidity and total suspended solids in ponds with water exchange near the end of the study. Energy costs were 31.5% higher for the ponds operated with water exchange than for the no-exchange ponds. KEY WORDS: Shrimp, ponds, water exchange INTRODUCTION Although shrimp fanning is a relatively benign activity com- pared with many other types of agriculture, industry, and residen- tial development, it may adversely affect coastal environments, especially where the regulatory framework is weak or inappropri- ate and shrimp farm development is intense. In the United States, where shrimp farms are widely dispersed and operate within rigid environmental protection guidelines, it is unlikely that significant environmental effects will occur. However, this is not the case in many other areas, where production collapses have often accom- panied intensive, poorly regulated development (e.g.. Taiwan [Liao 1992], Thailand [Lin 1995], China [Wang et al. 1995]). The effects are felt by both the shrimp farmer and other users of public water resources. The potential environmental effects of shrimp farming include: (1) wetland destruction for the construction of shrimp ponds; (2) hypemutrification of estuarine ecosystems by shrimp pond efflu- ent; (3) "biological pollution" of native shrimp stocks through escapement of aquaculture stocks; 14) excessive water use and entrainment of estuarine biota by pumping; (5) release and spread of disease organisms; and (6) discharge of treatment chemicals into estuarine systems (Hopkins et al. 1995a). Of these, all but the first can be addressed directly through improved water manage- ment, particularly as it relates to routine water exchange and dis- charge. Discharges from shrimp ponds may occur as a result of: II) runoff from heavy precipitation or additions to maintain salinity during periods of high evaporation; (2) routine, intentional water exchange to dilute pond water column nutrients, solids, and bio- chemical oxygen demand (BOD) (Lee and Wickins 1992); and (3) drain harvesting (Hopkins et al. 1995b). The magnitude and effect of these categories vary considerably, but routine water exchange generates the largest volume of effluent and is the easiest to avoid. Water exchange practices are seldom based on nutrient moni- *Corresponding author. toring or other "hard" data or used in response to fluctuating environmental conditions. Instead, exchange is often based on a set schedule (Hopkins and Villalon 1992). In extensive production systems, water exchange is sometimes used to introduce additional forage prey (Allan and Maguire 1993) and/or recruit wild postlar- vae/juveniles (Whetstone et al. 1988). However, for the most part, systematic investigations to determine how much water exchange is actually needed are generally lacking (e.g., see Allan and Magu- ire 1993; Hopkins etal. 1993, Hopkins et al. 1995b. Hopkins et al. 1995c). Despite the widespread belief that regular water exchange will improve pond water quality, there is also potential for the oppo- site. For example, a shrimp farmer in South Carolina began ex- changing large amounts of water in reaction to slightly increased, but not lethal, water column ammonia concentrations (Waddell Mariculture Center IWMC] unpublished data). The rapid water exchange effectively flushed the phytoplankton population from the pond, but because feeding was not discontinued, it did not affect the production of ammonia through metabolic processes occurring on the pond bottom (e.g. . shrimp metabolism, microbial decay processes). Phytoplankton and nitrifying bacteria attached to detrital particles in the water column are probably the primary routes of water column ammonia recycling in pond systems. By removing these components, rapid water exchange subsequently increased the amount of free water column ammonia. Once water exchange was stopped or reduced, ammonia concentrations quickly climbed to very high and potentially toxic levels. Had massive water exchange not disrupted the system's ability to reach equilibrium at the feeding rates being used, the pond ecosystem should have been able to react to the increased ammonia concen- tration through increased primary production and denitrification. Only by reducing water exchange were phytoplankton and nitri- fying bacteria populations allowed to rebound and affect a long- term reduction in ammonia. Similariy. BOD is dictated by respiratory and decay processes at the pond bottom and in the water column. Sources of oxygen are atmospheric diffusion across the water surface and oxygen gener- ation by photosynthesis. Diffusion rates are a function of the con- 441 442 Hopkins et al. centration differential across the surface boundary and are en- iianced by the surface disturbance and mixing caused by wind or aeration equipment (Boyd 1990). With adequate sunlight, the ox- ygen released through photosynlheses during the day exceeds combined day and night phytoplankton respiration. The effect of water exchange on oxygen diffusion rates is min- imal, and diffusion can be increased more efficiently through the use of aeration equipment. Even when the inlet water is saturated with oxygen, the pumping head and pump efficiency generally make it less expensive to operate aerators than to pump water. The effect of water exchange on photosynthetic oxygen enrichment depends on the sunlight available to algae. Available sunlight is a function of solar radiation, pond depth, density of the phytoplank- ton population, and amount of other nonphytoplankton turbidity. The phytoplankton population can become so dense that it causes a self-shading effect where light is absorbed before it penetrates far into the water column. On very cloudy days and/or at times of very dense phytoplankton populations, the available sunlight does not induce enough photosynthetic oxygen production to exceed the 24-h algal respiration or produces an excess of oxygen that is too small to offset other respiratory demands. Thus, when solar radi- ation is low and algal populations are very dense, phytoplankton dilution through water exchange may increase light penetration and improve the overall oxygen balance. However, as in the case of ammonia dilution noted above, excessive flushing of phyto- plankton may increase the oxygen deficit if the diminished phy- toplankton population cannot effectively use the solar radiation for oxygen production. Ideally, the phytoplankton population would reach an equilibrium where its density is controlled by available sunlight at a level that stabilizes oxygen production at a level sufficient to meet all pond respiratory requirements. Discharges from intentional water exchange in shrimp farming can be substantial. The production of a metric ton of shrimp typ- ically uses 55,000-86,000 metric tons of water (Hopkins and Vil- lalon 1992). However, large differences in the reported water con- sumption between species and systems cannot be explained by interspecific differences in metabolism or physiological toler- ances. This suggests that there is room for much improvement in water management (Hopkins and Villalon 1992, Phillips et al. 1991). In the past, the amount of water exchanges has generally been increased with increasing intensity of production (although not in strict linear fashion), reaching 3:30%/day at very high stocking densities (Sandifer et al. 1991). In response to the need to reduce the potential for environmen- tal effects from shrimp farming, researchers at WMC in South Carolina began examining the importance of water exchange in 1990 (Hopkins et al. 1991. Hopkins et al. 1993. Hopkins et al. 1995c, Hopkins et al. 1995d; Browdy et al. 1993, Sandifer and Hopkins 1996). Those authors found that water exchange could be reduced to low levels (^4%/day) or eliminated altogether without negatively affecting shrimp production as long as adequate levels of dissolved oxygen (DO) were maintained (see Hopkins et al. 1995a and Hopkins et al. 1995b for review). However, these stud- ies generally lacked replication or were conducted only in tanks. This experiment was undertaken to provide a more systematic comparison of exchange and no-exchange water management strategies for the intensive culture of shrimp in earthen ponds. MATERIALS AND METHODS The study was conducted during 1995 at the WMC. a field station of the Marine Resources Division, South Carolina Depart- ment of Natural Resources. Three replicate 0.1-ha (1,008-m"), 1,300-m' ponds were used for each of two treatments: 0 and 15% of pond water volume exchanged daily. The treatments were in- stituted after an initial 49-d period of no water exchange immedi- ately after stocking. These ponds have a 2: 1 length:width ratio, 3:1 side slopes, and a sloped bottom where depth ranges from 1.3 to 1.5 m; they are lined with 1-mm-thick high-density polyethylene that is overlaid on the bottom with 26 cm of the sandy native soil. A concrete drain structure incorporates screen tracks, an overflow weir, and a drain valve. The ponds were designed to drain into an effluent ditch that carries water back to an adjacent estuary some 800 m from intake pumps. This estuary, the Colleton River, is a high-salinity, vertically mixed embayment. The ponds were filled to within 20 cm of the overflow structure 1 wk before stocking. As ponds were filled, water was screened through a mesh sock. Ponds were stocked with shrimp {.Penaeus vcininimei. Boone) postlarvae at a density of 38.2 animals/m" on April 11-12, 1995. The postlarvae were raised at a commercial shrimp hatchery in South Carolina from nauplii produced from specific-pathogen-free broodstock at WMC. Shrimp growth was monitored at 14-d intervals by the shrimp being captured and individually weighed to the nearest 0.1 g. a random sample of 100 shrimp. The ponds were harvested after 153 d, at which time the entire crop was weighed en masse and sub- samples of shrimp were weighed individually to estimate harvest size and survival. Production characteristics of the two treatments were compared statistically using a t-test. Ponds were fed once daily from a tractor-drawn pneumatic feed TABLE 1. Production characteristics of intensive shrimp ponds operated with and without water exchange. Treatment 15% /Day Water Exchange No Water Exchange Pond/replicate designation Replicate mean harvest weight (g) Treatment mean harvest weight (g) Replicate survival (%) Treatment mean survival (%) Replicate production (kg/ha per crop) Treatment mean production Peplicale feed conversion efficiency Treatmeni mean FCR S-06 S-07 S-08 S-10 S-11 S-12 15.4 17.5 16.5 16.7 15.5 15.1 15.6 16.3 93,8 91.3 93.4 95.1 87.0 90.6 91.2 96.1 5.510 6,096 5,888 6,058 5.134 5,233 5,444 5.965 1.58 1.43 1 48 1,44 1.70 1.66 1.60 1.46 Effect of Water Exchange on Shrimp Production 443 TABLE 2. Average of weekly water quality analyses for each pond at the heginning of the study, the treatment mean, and the number of samples (n) Each \ alue Represents. Treatment Pond pH TURB SAL BOD TAN PHOS TSS With exchange S-06 8.2 6.3 26.0 21.5 0.15 1.18 103.3 S-07 8 3 4.0 25.8 17.2 0.86 0.89 69.5 S-08 8 T 3.0 25.5 17.6 0.82 099 74.2 Mean 8 -) 4.4 25.8 18.8 0.61 1.02 82.3 Without exchange S-10 8 2 2.8 25.3 14.8 0.42 1.58 88.7 S-Il 8 3 3.8 25.3 16.8 0.76 1.17 82.8 s-i: 8 2 2.7 25.3 15.1 0.64 1.14 83.8 Mean 8 •> 3,1 25.3 15.6 0.61 1 29 85.1 n 4 4 4 3 -I 1 3 TURB. nephelometer turhidity; S.^L, salinity, TAN. total ammonia nitrogen. PHOS. reactive orthophosphate; TSS. total suspended solids. blower. The feed was a 40% protein, 3-mm-diameter x 12-mm pellet manufactured for shrimp by Rangen Inc. (Buhl ID), In con- trast to typical feeding procedures where the amount of feed pre- sented varies with shrimp size and estimated standing crop bio- mass (e.g., see Clifford 1985). a constant daily feed amount of 5.67 kg/pond per day (57 kg/ha per day) was used in this study. This approach minimizes fluctuations in organic loading due to feeding in an effort to stabilize chemical and microbial processes in the ponds. Each pond was aerated continuously with a modified l-hp ■"Taiwanese-style"' paddlewheel aerator. A single paddlewheel provided enough supplemental aeration to maintain dawn dis- solved oxygen concentrations within an acceptable range. How- ever, to ensure that an aerator malfunction did not result in dawn dissolved oxygen depletion, a second l-hp paddlewheel was turned on and off automatically by a time switch between 0100 and 0700 h each day. The total aeration rate was, thus, 12.5 hp-day/ha per day. Aerators were situated in opposite comers of ponds so as to create a gyre that effectively swept most of the pond bottom and deposited sludge particles in the center. Power consumption and electrical costs for pumping and aerating water were monitored throughout the study. Permit conditions for raising nonindigenous shrimp species in South Carolina require that discharge structures be double screened and that there be no discharge until the shrimp reach a mean size of 1 g. This stipulation is designed to prevent postlarvae and small juveniles from moving around the screen frame and escaping to the estuary via the discharged water. Therefore, water exchange was begun in three of the six ponds on day 49. after the first bi-weekly sample where the shrimp size averaged more than 1 g. The water exchange rate was adjusted to 15% of the pond volume per day with a calibrated weir tube. The flow rate was checked daily and adjusted as necessary. The three no-exchange ponds received no additional water except that contributed by pre- cipitation. During August, heavy rains generated an estimated dis- charge of 0.6% of pond volume/day or about 0. 15%/day over the production cycle. Temperature and DO were measured daily at dawn by therm- istor and polarographic meter. We intended to measure a variety of other water-quality parameters as well, but the Center's water- quality chemist became seriously ill and was unable to conduct those analyses. Near the end of the study, a graduate student determined the pH. salinity, nephelometer turbidity, BOD. total ammonia nitrogen, reactive orthophosphate, and total suspended solids three times at roughly weekly intervals for each pond using standard methods (APHA 1989). RESULTS AND DISCUSSION Mean shrimp size at harvest size ranged from 15.1 to 16.7 g, and there was considerable overlap between treatments (Table 1). The overall mean harvest size for all ponds with water exchange was 0.9 g greater than that of ponds without water exchange, but this difference was not statistically significant when compared TABLE 3. Average of weekly water quality analyses for each pond near the end of the study, the treatment mean, and the number of samples (n) each value represents. Treatment Pond pH TURB SAL BOD TAN PHOS TSS With exchange S-06 7.7 28.7 21.0 26.31 0.89 0.17 57.5 S-07 7.8 29.3 22.2 13.43 1.35 0.05 53.0 S-08 8.1 23,0 22.7 14.67 0.51 0.01 97.5 Mean 7.9 27.0 21.9 18.14 0.92 0.08 69.3 Without exchange S-10 7.9 14.7 19.5 42.81 3.50 0.38 44.5 S-II 7.9 18.7 18.8 39.00 2.95 0.57 46.0 S-12 7.8 14.3 17.0 36,65 3.02 0.33 38.5 Mean 7.9 15.9 18.4 39.49 3.16 0.42 43.0 n = 3 3 3 3 3 3 2 See footnote to Table 2 for explanation of abbreviations. 444 Hopkins et al. TABLE 4. Estimated nutrient load discharged from ponds with and without routine water exchange on a continuing daily basis during the final 2 wk of the study and during the drain-harvesting process. All values are expressed as kg/ha. Discharge BOD TAN PHOS TSS Daily Exchange ponds 35.4 1.8 0.2 135.2 No-exchange ponds Trace Trace Trace Trace One-Time Harvest Exchange ponds 235.8 11.9 1.0 901.3 No-exchange ponds 513.3 41.0 5.5 559.0 All values are expressed as kilograms per hectare. See footnote to Table 2 for explanation of abbreviations. with a t-test (P = 0.2637). The size distributions were quite sim- ilar for both treatments, and all pond shrimp populations fell into the same market category (26-30 whole shrimp per pound) with the same value ($3.75/kg whole weight). Survival in all ponds was excellent, ranging from 87 to 96% . Again, although the average survival in the no-exchange ponds was slightly lower than that for the exchange ponds (Table 1 ). this difference was not significant when examined by t-test (P = 0.4946). Production, a function of growth and survival, ranged from 5,134 to 6,096 kg/ha per crop whole weight, with considerable overlap between treatments. Average production for the exchange and no-exchange treatments was 5,888 and 5,444 kg/ha per crop, respectively, and this differences was not significant (P = 0.2416). Because all ponds received the same feed mput, the feed con- version ratio (FCR) (kilograms of dry feed;kilograms of whole shrimp) was also a function of growth and survival and ranged from 1.43:1 (apparently the lowest yet reported for intensive pond culture of shrimp) to 1 .69; 1 . A t-test found no statistically signif- icant difference in FCR between treatments (P = 0.2395). For all ponds and all days of the 153-d study, DO averaged 5.15 and 5.04 mg/1 for the water exchange and no-exchange treat- ments, respectively. This difference was significant (P = 0.0007). Excluding the first 49 d. when there was no water exchange in either treatment, DO averaged 4.92 and 4.73 mg/l, respectively, for the exchange and no-exchange treatments. The minimum dawn DO for all days and all ponds was 2.5 mg/l. At the beginning of the study, indicators of water quality were similar in all ponds, although turbidity and BOD tended to be slightly higher in the exchange ponds (Table 2). Near the end of the experiment, when pond water quality was sampled at weekly intervals, there was more variation in pH within than between treatments (Table 3). By the end of the study, salinity was 3—^ ppt lower in the no-exchange ponds because of precipitation. The Colleton River estuary has naturally high turbidity and suspended solids, especially after heavy rains. Therefore, water in the ponds with continuous water exchange had higher nephelometer turbidity and total suspended solids than did water in the no-exchange ponds. As expected, total ammonia nitrogen, reactive orthophos- phate, and 5-d BOD were higher in the no-exchange ponds than in the ponds with water exchange (Table 3). Actual electricity costs at WMC, where industrial rates apply, averaged $0.058/kwh (range, $0.053-$0.070/kwh) over the course of the study. The electrical efficiency of the aerators was 1 .3 kwh/h per rated horsepower. With an average aeration rate of 12.5 hp-h/h per hectare, the electrical cost for aeration over the 153-d study was $3.461/ha per crop for all ponds. WMC pumps water from the estuary to the ponds with 30-hp centrifugal pumps. The efficiency of the pumps is 7.24 m^^/kwh. All ponds required $89/ha per crop for electricity to fill them initially. By exchanging water at a rate of 15% of the pond volume per day for 104 d of the study, the ponds receiving water exchange had an additional pumping cost of $1 ,634/ha per crop. Therefore, the ponds receiv- ing water exchange and the no-exchange ponds had electrical pumping costs of $1,723 and S89/ha per crop, respectively. The total cost for electricity (aeration + pumping) was $5,184/ha for the exchange ponds vs. $3,550/ha for the no-exchange treatment in this study. Thus, for the WMC system, the no-exchange water management strategy resulted in a 31.5% savings in electrical costs. In the United States, the Environmental Protection Agency (EPA) and/or the state agency authorized to administer EPA reg- ulations typically requires that a National Pollution Discharge Elimination System (NPDES) permit be issued to warmwater aquaculture facilities that produce more than 45.4 metric tons ( 100,000 pounds) of product per year and discharge water more than 30 d/year. With production goals similar to those used in this study (i.e., 5,()0()-6.000 kg/ha per crop), a shrimp fami with 8 ha or more of ponds operated with routine water exchange would likely require a NPDES permit. On the basis of our experience with requirements of NPDES permits for aquaculture operations in South Carolina (which may not be representative of other states or regions), we estimate compliance costs at approximately $400/ha per year for an 8-ha intensive shrimp farm. In addition, farms designed for no-exchange technology could reduce the scope and costs of water pumping and distribution systems. The more important cost associated with water exchange is the environmental cost. In regions with poorly regulated shrimp- farming activity, the environmental cost may include a collapse of the shrimp-farming industry or disruption of other public uses of coastal waters. In countries like the United States, where there is a more comprehensive regulatory structure, industry expansion may be prohibited if all farms use routine water exchange. The no-exchange ponds discharged more BOD, ammonia ni- trogen, and phosphorus during the drain-harvesting process than did the ponds with water exchange, but the ponds with water exchange released more total suspended solids (Table 4). As noted by Hopkins et al. (1993), the total load of nutrients discharged through the production process (water exchange plus drain har- vest) is much higher when routine water exchange is used. As can be seen from Table 4. the differences in drain harvest discharge load between ponds operated with and without water exchange would be overshadowed by the daily discharge associated with routine water exchange. Without routine water exchange, in situ assimilation and digestion processes mineralize and deposit much of the nutrient mass in pond bottom sediments or volatilize them to the atmosphere. ACKNOWLEDGMENTS This is contribution No. 370 from the Marine Resources Divi- sion of the South Carolina Department of Natural Resources. Mr. Peter Hamilton and Mr. Chris Caldwell are thanked for their care- ful attention to the daily management of the study. Ms. Patricia Kinne did water quality analysis in the final weeks of the study, and Mr. David Smith provided climatological data. We will miss Effect of Water Exchange on Shrimp Production 445 the continued contribution of the late Mr. Joseph Hoats, WMC biologist and chemist. This study was conducted as part of the Gulf Coast Research Laboratory Consortium's U.S. Marine Shrimp Farming Program and was tunded, in part, by the U.S. Department of Agriculture, Cooperative States Research, Educa- tion and Extension Service, via a subcontract from the Oceanic Institute. LITERATURE CITED Allan, G. L. & G. B. Maguire. 1993. The etfecl of water exchange on production oi Mehipenueiis macleayi and water quality in experimental ponds, y. World Aiimuullure Soc. 24(3):321-.^28. APHA (American Public Health Association!, American Water Works Association, and Water Pollution Control Federation. 1989. Standard Methods for the Examination of Water and Wastewater. 17th ed. American Public Health Association, Washington, DC. Boyd. C. E. 1990. Water Quality in Ponds for Aquaculture. Alabama Agricultural Experiment Station. Aubum University. Alabama. Browdy. C. L.. J. D. Holloway. C. O. King. A. D. Stokes. J. S. Hop- kins & P. A. Sandifer. 1993. IHHN virus and intensive culture of Penaeus vanimmei: effects of stocking density and water exchange rates. J. Cntslaceait Biol. I3(l):87-94. Clifford. H. C. 1985. Semi-intensive shrimp fanning, pp IV-IS lo lV-42. In: G. A. Chamberlain, M. H. Haby and R. J. Miget (eds.). Texas Shrimp Fanning Manual. Texas Agricultural Extension Service, Cor- pus Chnstj, Texas. Hopkins, J. S., C. L. Browdy, R. D, Hamilton 11 & J. A. Heffeman 111. 1995c. The effect of low-rate sand filtration and modified feed man- agement on effluent quality, pond water quality and production of intensive shrimp ponds. Esmaries 18( 1A):1 16-123. Hopkins, J. S., C. L. Browdy, P. A. Sandifer & A. D. Stokes. 1995d Effect of two feed protein levels and feed rate combinations on water quality and production of intensive shrimp ponds operated without water exchange. J. World Aquaculture Soc. 26(l):93-97. Hopkins, J S., R. D. Hamilton. P. A. Sandifer. C. L. Browdy & A. D. Stokes. 1993. Effect of water exchange rates on production, water quality, effluent characteristics and nitrogen budgets in intensive shrimp ponds. J. World Aquaculture Soc. 24(3):304-320. Hopkms, J. S., P. A. Sandifer & C. L. Browdy. 1995b. A review of water management regimes which abate the environmental impacts of shnmp fanning, pp. 157-166. In: C. L. Browdy and J. S. Hopkins (eds.). Swimming Through Troubled Water, Proceedings of the Spe- cial Session on Shrimp Farming, Aquaculture '95. World Aquaculture Society, Baton Rouge. Louisiana. Hopkins, J. S., P. A. Sandifer. M. R. DeVoe, A. F. Holland. C. L. Browdy & A. D. Stokes. 1995a. The environmental impacts of shrimp farming with special reference to the situation in the continental U.S. Estuaries I8( lA);25-42. Hopkins. J. S . P. A. Sandifer. A. D Stokes & C. L. Browdy 1991 The effect of minimal water exchange on the water quality and production of intensive marine shrimp ponds, p. 33. Program and Abstracts. 22nd Annual Conference. World Aquaculture Society. WAS. Baton Rouge. Louisiana. Hopkins. J. S. & J. Villalon. 1992. Synopsis of industrial panel input on shrimp pond management, pp. 138-143. In: J. A. Wyban (ed.). Pro- ceedings of the Special Session on Shrimp Farming. World Aquacul- ture Society. Baton Rouge. LA. Lee. D. O'C. & J, F. Wickins. 1992, Crustacean Farming. Blackwell Scientific Publications. Oxford. United Kingdom. Liao. 1-C. 1992. Marine prawn culture industry of Taiwan, pp. 653-675. In: A. W. Fast and L. J. Lester (eds.). Marine Shrimp Culture: Pnn- ciples and Practices. Elsevier Science Publishers B. V.. Amsterdam. Lin. C K 1995. Progression of intensive marine shrimp culture in Thai- land, pp. 13-23. In: C. L. Browdy and J. S. Hopkins (eds.). Swim- ming Through Troubled Water. Proceedings of the Special Session on Shrimp Farming. Aquaculture '95. World Aquaculture Society. Baton Rouge. Louisiana. Phillips, M. J.. M. C. M. Beveridge & R. M. Clarke. 1991. Impact of aquaculture on water resources, pp. 568-591. In: D. E. Brune and J. R. Tomasso (eds). Aquaculture and Water Quality. Advances in World Aquaculture- vol. 3. Worid Aquaculture Society. Baton Rouge. Louisiana. Sandifer. P. A.&J S.Hopkins. 1996. Conceptual design of a sustainable pond-based shrimp culture system. Aquacultural Engineering 15(1): 41-52. .Sandifer. P. A.. A. D. Stokes. J. S. Hopkins & R. A. Smiley. 1991. Further intensification of pond shnmp culture in South Carolina, pp. 84-95. In: P. A. Sandifer (ed). Shrimp Culture in North America and the Caribbean. World Aquaculture Society. Baton Rouge. Louisiana. Wang. Q.. Y. Conghai & Y. Jia. 1995. The shrimp farming industry in China: Past Development, present status and perspectives on the fu- ture, pp. I-I2. /«. C. L. Browdy and J. S. Hopkins (eds.). Swimming Through Troubled Water. Proceedings of the Special Session on Shrimp Farming. Aquaculture '95. World AquaculUire Society. Baton Rouge. Louisiana. Whetstone. J. M.. E. J. Olmi III & P. A. Sandifer. 1988. Management of existing saltmarsh impoundments in South Carolina for shrimp aqua- culture and its implications, pp. 327-338. In: D. D. Hook et al. (eds.). The Ecology and Management of Wetlands, vol. 2. Management and Value of Wetlands. Croom Helm. London: Timber Press. Portland. Oregon. 394 pp. Journal of Shellfish Research. Vo\. 15, No. 2. 447-464. 19%. ABSTRACTS OF TECHNICAL PAPERS Presented at the 16th Annual Meeting MILFORD AQUACULTURE SEMINAR Milford, Connecticut February 26-28, 1996 447 Milford Aquacullure Seminar. Milford. Connecticut Abstracts. February 26-28, 1996 449 CONTENTS Walter J. Blogoslawski Overview. 16th Milford Aquaculturc Seminar 451 Jennifer H. AILx. Mark S. Dixon. Barry C. Smith, and Gary H. Wikfors Scallop larval feeding experiments: some surprises and unanswered questions 45 1 Joseph Chorotnanski, Sheila Stiles, Daniel Schweitzer, Matthew Mroczka, and Paul Dinwoodie Comparison of three Long Island Sound sites for the grow-out of bay scallops, Argopecten irrudians 451 John Curtis A report on the project for suspension culture of bay scallops in Long Island Sound 45 1 Mark S. Dixon. Jennifer H. Alix, Barry C. Smith, and Gary H. Wikfors Division rates of five high-lipid Telrasehnis strains at temperatures from 10°C to 35°C 452 Judy Dutra Truro sea scallop aquaculturc project 452 Frank A. Dutra, Scott C. Feindel, and Robert Garrison A novel technique for field deployment for post-set Argopeclen irradinns: avoiding the Jiffy-Pop Syndrome 453 C. Austin Farley, Farl J. Lewis, David Relyea, Joseph Zahtila, and Gregg Rivara Resistance studies for juvenile oyster disease (JOD): 1995 continuation 453 C. Austin Farley, Roy Scott, and Leon Williams Resistance studies of Chesapeake Bay. Maryland, oysters: epizootiology of two populations exposed to HaplosporiJium nelsoni and Perkinsus inurimis 454 Alexander Gryska, G. Jay Parsons, Sandra E. Shumway, Kristin P. Geib, Ian Emery, and Sue Kuenstner Polyculture of sea scallops suspended from salmon cages 454 Bart Harrison, Karin A. Tammi, and Wayne H. Turner Algal fouling and predation of artificial spat collectors in the Westport River Estuary, Massachusetts 454 Porter Hoagland and Di Jin Designing an access system for ocean mariculture 455 Hunt Howell, Linda KUng, Terry Bradley, and Larry Buckley Development of a groundfish aquaculturc industry in New England: review of efforts to date 455 Richard C. Karney Hatchery culture techniques for the giant sea scallop Placopccten nuigellaniciis 455 John Kraeuter, John Aldred, Paul Bagnall, Richard Crema, Gef Flimlin, Susan Ford, Dale Leavitt, George Mathis, Gregg Rivara, Roxanna Smolowilz, and Margy Wintennyer Overview of seed hard clam winter mortality studies in New Jersey, New York and Massachusetts 456 J. W. Latchford, S. B. Prayitno, and A. Alabi The use of vaccines in the culture of penaeid prawns 456 Mijin Lee, Gordon T. Taylor, Monica Bricelj, and Susan E. Ford Experimental evaluation of Vibrio spp. as etiological agent of JOD 456 Earl J. Lewis, C. Austin Farley, Ana Baya and Rosangela Navarro 1995 juvenile oyster disease (JOD) transmission and bacteriological studies 457 Shawn M. McLaughlin Diagnosis and prevalence of Perkinsus sp. in Chesapeake Bay softshell clams, Mya Arenaria: an update 457 Harold C. Mears Fig's and aquaculturc — a look back and a look ahead 458 Matthew Mroczka, Paul Dinwoodie, Ronald Goldberg, Jose Pereira, Paul Clark, Sheila Stiles, Joseph Choromanski, Daniel Schweitzer and Nancy Balcom Culture of the bay scallop, Argopecten irradians. within a small-boat marina on Long Island Sound (Connecticut) 458 George Nardi Culture of summer flounder. Paralichthys dentatus. at GreatBay Aquafarms 458 Michael J. Oesterling and Laura A. Rose Bay scallop culture in a Virginia saltwater pond 458 Dean M. Perry, Renee Mercaldo- Allen, Catherine Kuropat, and James Hughes Laboratory culture of lautog: a pilot study 459 Steven Pitchford A simple system for long-term exposures of adult or larval bivalves to bacterial pathogens 459 450 Abstracts. February 26-28, 1996 Milford Aquaculture Seminar, Milford, Connecticut Walter Y. Rhee Lessons from scallop spat collection in Washington state 459 Michael A. Rice and Joseph T. DeAlteris Changing times for Rhode Island shellfisheries; URFs shellfish aquaculture training program for shellfishermen 459 Elizabeth F. Scolten, Gabriella C. Castro, and Debra L. Colombo Martha's Vineyard shellfish group aquaculture training program 460 Barry C. Smith, Gary H. Wikfors, Jennifer H. Alix, and Mark S. Dixon A PC-controlled, expermiental moUuscan rearing apparatus for studies of feeding strategies and nutrition 460 Roxanna Smolowitz, Dale Leavilt, and Frank Perkins An important new disease of hard clams, Mercenaria merceiuiria. in the northeast United States 460 Sheila Stiles, Joseph Choromanski, Daniel Schweitzer, and Qin-Zhao Xue fteliminary investigations of genetics and breeding of the bay scallop, Argopecten irradians 46 1 Karin A. Tammi, Wayne H. Turner, Margaret Brumsted, and Michael A. Rice Veliger vodoo and other witch craft in the Westport River: bay scallop, Argopecten irradians. veliger abundance and the variability of spatfall recruitment to artificial spat collectors in the Westport Estuary. Massachusetts 461 Eric M. Thunberg Update on federal policy affecting marine aquaculture in the exclusive economic zone of the United States 462 Wayne H. Turner, Karin A. Tammi, Bart Harrison, and Bethany A. Starr What a year to be a mud crab! Three years on the bay scallop restoration project, Westport River Estuary. Massachusetts 462 James C. Widman, Jr. and Christopher G. Cooper Growth of bay scallops, Argopecten irradians. in 5 mm mesh lantern nets 463 Gary H. Wikfors, Barry C. Smith, Jennifer H. Alix, and Mark S. Dixon Feeding strategies for post-set bay scallops, Argopecten irradians: What? How Much^" How Often? 463 Paul Willoughby and Jack Blake Pioneering efforts to privately culture quahogs, Mercenaria mercenaria, m the town of Edgartown, MA 463 Frederick B. Wishner Murphy's law and the raising of atlantic sturgeon 464 Milford Aquaculture Seminar. Milford. Connecticut Abstracts. February 26-28. 1996 451 OVERVIEW, 16TH MILFORD AQUACULTURE SEMI- NAR. Walter Blogoslawski, U.S. Department of Commerce. Na- tional Oceanic and Atmospheric Administration, National Marine Fisheries Service. Northeast Fisheries Science Center, 212 Rogers Avenue. Milford. CT 06460. The 16th Annual Milford Aquaculture Seminar attracted 37 eminently qualified speakers, whose topics ranged from shellfish and finfish disease and propagation through sponsorship of spe- cific programs designed to assist displaced fisherman in the de- velopment of aquaculture ventures. The 160 attendees from the United States, Canada, the United Kingdom and the People's Re- public of China met in formal and informal sessions to discuss the latest developments in technology and government extension ac- tivities. It was noted with interest by attendees that many of the concerns of the industry, such as overharvesting in the late 19th century, were the same as those influencing shellfish farmers to- day. Speakers from 12 states, Canada, and the United Kingdom discussed sea and bay scallop propagation and the control of shell- fish diseases as well as aquaculture trammg and culture of finfish. Twenty-six marine laboratories and hatcheries and twenty univer- sities, were represented during this exchange. Genetics, nutrition, and disease, as related to the culture and survival of shellfish, were prominent topics of discussion. Sponsorship of this seminar included the National Marine Fish- eries Services. Milford Laboratory. Milford. Connecticut and the United States Department of Agriculture. Northeastern Regional Aquaculture Center in North Dartmouth. Massachusetts. Their support is most gratefully acknowledged. SCALLOP LARVAL FEEDING EXPERIMENTS: SOME SURPRISES AND UNANSWERED QUESTIONS. Jennifer H. Alix,' Mark S. Dixon, '^ Barry C. Smith,' and Gary H. Wikfors,' 'USDOC. NCAA, National Marine Fisheries Ser- vice, Northeast Fisheries Science Center, Milford, CT 06460: ^Marine Sciences and Technology Center. University of Connect- icut. Groton. CT 06340. Controlled feeding studies with larval bay scallops (Argopecten irradians) are being pursued with an ultimate goal of reducing larval rearing to seven days or fewer as a convenience for hatchery work schedules. Initial experiments compared larval scallop survival. growth, and time to metamorphosis when fed unialgal diets from a wide variety of microalgal taxa. Cells larger than about 7-8 |xm were too large for first-feeding larvae, and a number of small chlorophytes and eustigmatophytes with cellulosic cell walls were indigestible. Two diatoms used widely as larval diets for other bivalves. Chaeto- ceros cakitrans and Thalassiosira pseudonana. were poor relative to a number of chrysophytes and prymnesiophytes. Two rarely-cultured strains of the flagellate. Pavlova. CCMP459 and CCMP609. sup- ported particularly good survival and growth; these strains contain remarkably high levels of essential fatty acid. Subsequent experiments investigated mixed algal diets, either from first feeding or with sequential replacement during larval life. Mixed diets including CCMP459 supported larval growth equiv- alent to several other mixed diets, but consistently resulted in early metamorphosis m eight days at a smaller size of 140-160 |j.m than mixed diets not including this alga, which resulted in metamor- phosis in. 10-14 days at sizes generally >190 ^.m. Further exper- iments will be required to determine if CCMP459 induces meta- morphosis tiirough some possible chemical mechanism. We also investigated the possibility of replacing a portion of a CCMP459 diet with high-lipid Tetraselmis strains on days 3, 5. or 7 of larval rearing. Two Tetraselmis strains with a size range of 12-15 txm were not ingested prior to day 5 or 6; however, a 10 |jLm Tetraselmis strain (PLAT-P) was consumed between days 3 and 4. Scallop larvae fed mixed CCMP459 + PLAT-P grew as well as larvae fed CCMP459 alone, but metamorphosed two days later at a larger size. Further work will be required to determine if early metamorphosis on diets including CCMP459 is an advantage or disadvantage with regard to subsequent postset performance. This project was funded partially by a University of Connect- icut. Marine Sciences and Technology Center Grant. COMPARISON OF THREE LONG ISLAND SOUND SITES FOR THE GROW-OUT OF BAY SCALLOPS, AR- GOPECTEN IRRADIANS. Joseph Choromanski,' Sheila Stiles,' Daniel Schweitzer,'" Matthew Mroczka,' and Paul Dinwoodie,-' 'USDOC. NCAA, National Marine Fisheries Ser- vice. Milford Laboratory. Milford CT 06460: "Marine Sciences and Technology Center. University of Connecticut. Groton. CT 06340; 'Cedar Island Marina, P.O. Box 181, Clinton. CT 06413. A preliminary study of growth and survival for selected strains of bay scallops was conducted at three sites in Long Island Sound. Float-supported lantern nets held off the bottom were utilized at off-shore sites in Groton and Milford. In Clinton, a suspension- culture rack system in a protected marina was used and evaluated for the first time. Various genetic lines of selectively-bred scallops were cultured from native Connecticut stock in our laboratory hatchery and held in raceways prior to deployment. Sampling was performed on a monthly basis, during which times we were able to assess conditions and equipment at the sites. This project was funded partially by a University of Connect- icut, Marine Sciences and Technology Center Grant. A REPORT ON THE PROJECT FOR SUSPENSION CUL- TURE OF BAY SCALLOPS IN LONG ISLAND SOUND. John Curtis, Bridgeport Regional Vocational Aquaculture School. 60 St. Stephens Road. Bridgeport. CT 06605. In December of 1994. the Bridgeport Regional Vocational Aquaculture School, in cooperation with the People's Republic of China, initiated a scallop restoration project using funds from the State of Connecticut's Long Island Sound License Plate Program 452 Abstracts. February 26-28. 1996 Milford Aquaculture Seminar, Milford, Connecticut to develop a small demonstration long-line farm in the waters off Fairfield, Connecticut. With the combined efforts of Dr. Luning Sun from Academia Sinica, and the students and staff of the Aqua- culture school, a complete scallop hatchery, as well as a grow-out farm, were constructed and maintained throughtout this past year. Students from the school grew algal cultures, spawned native scal- lops and made qualitative and quantitative growth measurements. Using proven scallop culturing techniques from China, the project participants successfully harvested more than 10,000 scallops this last December. The project has been continued m order to study various culture techniques and experimental factors and to inves- tigate its potential commercial feasibility in the area. This restoration project has proved to be successful both theo- retically and practically. It has involved regional high school stu- dents and provided the opportunity for them to work closely with the students and professors from several local universities, as well as the staff from various federal laboratories. sion rates were in the 0.5 div/day range, and reduced growth rates were less than 0. 1 div/day in cultures that were sluggish but still growing. These division rates are consistent with algae grown on a larger scale in carboys. The results of this experiment, coupled with previous work showing that MC:2. PLAT-P, and PLY429 support fast growth in scallops, can be used to design an effective feeding strategy. By growing each strain during the season of its most suitable temper- ature, heating and cooling costs associated with year-round, large- scale microalgal culture can be reduced. These results indicate that MC:2 and PLY429 would be good warm weather choices, and that PLAT-P would be a suitable cool weather choice. This project was funded partially by a University of Connect- icut. Marine Sciences and Technology Center Grant. DIVISION RATES OF FIVE HIGH-LIPID TETRASELMIS STRAINS AT TEMPERATURES FROM 10°C TO 35°C. Mark S. Dixon,'" Jennifer H. Alix.' Barry C. Smith.' and Gary H. Wikfors,' 'USDOC. NCAA, National Marine Fisheries Service, Northeast Fisheries Science Center, Milford, CT 06460; ^Marine Sciences and Technology Center. University of Connect- icut. Groton, CT 06340. Temperate climates present challenges to molluscan aquacul- ture; in spring, water and air temperatures are below optimal and in summer, above. Regardless of temperature conditions a contin- uous supply of quality microalgal food is an essential component of shellfish culture. The cost of temperature control can be less- ened by selecting algal strains that grow well at low or high tem- peratures to match the season. Five Tetraselmis strains were grown in test tubes under identical light conditions but at various tem- peratures (10°, 15°, 19°, 25°, 30°, and 35°C) in temperature- controlled incubators. Previous studies have shown that these five high-lipid Tetraselmis strains have excellent potential for maxi- mizing shellfish growth. Microalgal growth rates were recorded as divisions per day calculated from both direct cell counts and op- tical density measures. MC:2 {Tetraselmis sp.) and PLY429 {Tetraselmis chui) had greatest division rates when grown above 20°C, and showed re- duced growth rates below 20°C. PLY429 had reduced growth above 25°C, but MC:2 continued to divide rapidly up to 30°C. PLAT-P (Tetraselmis striata) and UW445 {Tetraselmis chui) both exhibited maximal growth rates below 20°C with a drop of divi- sion rate above 20°C. PLY429S {Tetraselmis chui) showed peak growth below 15°C and reduced growth above 15°C. Tempera- tures above 35°C were lethal to all strains within 10 days and resulted in negative growth in the short term. Typical peak divi- TRURO SEA SCALLOP AQUACULTURE PROJECT. Judy Dutra, Truro Aquaculture Project, 43 Shore Road, North Truro, MA 02652. The purpose of the Truro Aquaculture Project is to demonstrate the feasibility of cultured growth and development of the giant sea scallop (Placupecten magellanicus) in Cape Cod Bay, Massachu- setts. The 10 acre site is located 2 miles off Truro, MA in water 60-70 feet deep and, as a designated critical habitat for the North- em Right Whale, requires certain adaptations to minimize hazards and risk encounters for the whales. Working with National Marine Fisheries Service, the U.S. Army Corps of Engineers, and Cape Cod Resources, the facility design has eliminated all vertical lines, all hanging mid-water gear and "spar buoys" have replaced tradi- tional buoys. Plans for monitoring, emergencies and entangle- ments have been developed. Spat have been produced by the Martha's Vineyard Shellfish Group for grow-out at the Truro site. Spat collecting in the wild, using traditional methods, is also being investigated. Grow-out will consist of bottom enhancement, bottom caged culture and a bottom technique in which scallop spat is glued to ribbons of plastic mesh. The project director is a commercial fishermen and has expe- rience and expertise in handling heavy gear in deep water. His understanding of structures in seawater and repair of marine struc- tures are important assets for a project of this type. Underwater video and still photography are being used to document and mon- itor the facility as well as the habitat and animals. The restrictions placed upon the Massachusetts fisherman and the impact of the industry on stocks have lead to the development of alternative methods of shellfish and pelagic harvesting. Sea scallop aquaculture has the potential to prove an economically sound business venture thereby attracting interest from the local fishing community. Sea scallop ranching and fanning is one op- tion for an industry in need of change. Milford Aquaculture Seminar. Milford. Connecticut Abstracts. Fcbruarv 26-28. 1996 453 A NOVEL TECHNIQUE FOR FIELD DEPLOYMENT OF POST-SET ARGOPECTEN IRRADIANS: AVOIDING THE JIFFY-POP SYNDROME. Frank A. Dutra, Scott C. Feindel. and Robert Garrison, Nantuclcet Research and Education Foun- dation. 0 Easton Street. Nantucket, MA 02554. Rapid growth rate makes the bay scallop, Argopecten irradi- ans. a prime candidate species for aquaculture. During the nursery phase, however, this factor can become a constraint for all but the largest facilities. An abrupt weekly doubling and occasional tri- pling of volume can become problematic as optimal stocking sizes are attained and juvenile scallops overtlow existing nursery sys- tems. To avoid overcrowding, labor forces are severely taxed in an attempt to thin and deploy large numbers of scallops in a relatively brief time span. Deployment techniques that rely on fine mesh for post settlement retention are susceptible to reduced flow as a result of site-dependent particulates and fouling organisms. Increased mortality, stress, and reduced growth rates can result. Remote sites lacking direct access to hatchery facilities may find it diffi- cult or impossible to employ these techniques. In addition, sav- ings gained through an abbreviated hatchery and nursery period can be over-shadowed in the subsequent painstaking and labor intensive grading and washing processes. The inability of many growers to employ current post-set technologies has resulted in a wide array of field and short-based nursery sys- tems, including up-wellers. raceways, and floating nursery trays. In an attempt to lower overall production cost and minimize constraints of these existing nursery systems, experimental post- set stocking techniques were developed as part of Nantucket's NOAA-sponsored aquaculture program. Equal volumes of 3- week-old (400-800 |x) post-set scallops were provided with a variety of substrata in hatchery down-welling trays and left over- night for byssal attachment. Substrate sections with attached scal- lops were then deployed directly to the field in 2.5-mm mesh pearl nets. Densities were determined from randomly selected sections under a dissecting microscope. Anticipated losses due to the in- advertent release of 400-800-jj. spat through 2.5-mm openings were minimized by an apparent gregarious settlement preference, as reported for other bivalve species. Errant scallops were ob- served attached to the outsides of nets in randomly-distributed, oval shaped clusters. Predation and further detachment appeared to be negligible as scallops matured and were removed subsequently from the outsides of the nets after 3 weeks (5-10 mm). Overall retention was estimated to be greater than 70% . Growth rates of post-set scallops substantially exceeded those of nursery-reared cohorts, presumably due to an increase in available food supplies and a minimization of stress associated with repeated handling and overcrowding. Mortality and deformity rates were negligible throughout the trials. Market-sized scallops (35—40 mm) were ev- ident (10% of total) within 90 days of spawning. This technique has proved feasible at any stage of development from eyed larvae up to a 2-mm size range, thus effectively extending the stocking period of any particular batch of scallops, and thereby minimizing the peak labor loads associated with stocking on demand. It should be noted that further study and refinements of this post-set tech- nology are needed, including the feasibility of shipment from hatcheries to remote sites. The initial results may well be site- dependent and of a temporal nature. RESISTANCE STUDIES FOR JUVENILE OYSTER DIS- EASE (JOD); 1995 CONTINUATION. C. Austin Farley,' Earl J. Lewis,' David Relyea," Joseph Zahtila,' and Gregg Rivara,"^ 'USDOC, NCAA. National Marine Fisheries Service. Oxford, MD 21654-9724; "Frank M. Flower Co.. Oyster Bay. NY 1 1771; 'Cornell University. Cooperative Extension. Southold. NY 11971. The F| progeny from 1991 brood stocks selected on the basis of survival of exposure to juvenile oyster disease (JOD) were ex- posed to the disease at two infective sites in 1994 along with susceptible seed (FCT) from naive Connecticut brood stocks. In this study, the presumed resistant F, population demonstrated up to 7 times better survival. In the 2nd year of this study, susceptible progeny. F, resistant progeny, and F, progeny from 1993 F, generation brood stocks were produced from June 1995 spawnings at the Frank M. Flower hatchery in Bayville. NY. Seed were placed in nursery raft trays on August 4. 1995 and deployed at 7 sites in the Long Island area on August 28; (1) Oyster Bay. LI Sound; (2) Mattituck Creek, LI Sound; (3) Cedar Beach (SCMELC), Peconic Bay (PB); (4) Par- rino Pond, PB; (5) Parrino Lease, PB; (6) Grothe Lease, PB; and (7) Pickerell Lease, PB. Evaluations for JOD (size, shell checks, conchiolin prevalence in live and dead oysters, and mortality) were made from weekly or biweekly samples between August 28 and November 6, 1995. No mortality was seen on August 28. 1995. at the Flower site. Size-culled FCT susceptible runts (16-20 mm) had mortalities of 75% by October 30. 1995. Size-culled resistant F, and F, runts had mortalities of 0 to 3% during this time. UncuUed FCT seed oysters had a maximum mortality of 15% by October 30, while the F, and F, seed had mortalities of <2% over the same time period. At site 2. unculled FCT seed had mortalities of 72% by October 30, 1995. The F, and F, seed had mortalities of 90%. After one year, shell heights were about 49 and 57 mm, wet adductor muscle weights were 2.8 and 4.5 g. and growth rates were 0. 1 1 and 0. 13 mm per day. These growth rates were comparable to sea scallops cultured in Atlantic Canada. Reduced rates of survival were found during the latter part of the experiment and were attributable, in part, to heavy fouling by blue mussels. The potential for supple- mental income, diversification of the salmon aquaculture industry, and logistics of culturing scallops in conjunction with salmon will be discussed. ALGAL FOULING AND PREDATION OF ARTIFICIAL SPAT COLLECTORS IN THE WESTPORT RIVER ESTU- ARY, MASSACHUSETTS. Bart Harrison.' Karin A. Tammi," and Wayne H. Turner,' 'Water Works Group, P.O. Box 197, Westport Point, MA 02791; "Department of Fisheries, Animal and Veterinary Science, University of Rhode Island, Kingston, RI 02881. In 1993, the Bay Scallop Restoration Project (BSRP) was es- tablished with a focus on returning the once profitable and rich scallop fishery to the Westport River with the use of artificial spat collectors and spawner sanctuaries. Researchers have observed that spat collectors (2- to 4-mm plastic mesh onion bags containing monofilament) frequently become fouled by various organ- isms, especially algae and sediment. Typically, both brown {Sphacelaria sp) and green [Enteromorpha sp and Cladophora sp.) algal species foul the exterior of the bag. reducing the effec- tiveness of the spat collector. Beginning in June 1995, multiple longlines containing 20 spat collectors were deployed weekly to each of 5 study areas. After soaking for approximately 3 weeks, 20 spat collectors were harvested to assess predation and fouling. Spat-collector fouling was recorded on a scale from I to 5, with 5 being heavily fouled. Researchers continued to visit each study area throughout the summer and fall at 3-week intervals to record increasing fouling levels. Additionally, at Corey's Island, fouling and predation also was compared to assess the pert'ormance be- tween the original onion bag collector with that of a new com- mercial fine-mesh collector. After soaking for 3 weeks, all study areas had low fouling ratings (1). With the exception of Corey's Milford Aquacullure Seminar, Milford. Connecticut Abstracts. February 26-28. 1996 455 Island, fouling indices for all study areas remained low (between 1 to 2) even after six weeks of soaking. At Corey's Island, fouling of onion-bag collectors remained low (1 to 2 rating), but the fine- mesh collector displayed a greater degree of fouling and sedimen- tation (3 rating), most likely due to the smaller mesh size of the collector. In general, fouling indices gradually increased between the fourth and sixth study week, probably resulting from increased water temperature and influx of nutrients into the estuary. Long- lines harvested in the fall were heavily fouled by Enteromorpha. Sphacelaria and CUidophora. In addition, these collectors were fully colonized by sea squirts. Mogiila sp., and mud crabs. Pan- opeus sp.. with the latter being the primary predator. Scallop recruitment over the summer was unaffected by the algal fouling to the collectors. However, as fouling increases, it may make the collector an inhospitable environment by restricting flow through the bag and depleting nutrients, thereby hindering the growth of scallops within the collector. In addition, if there is a second spawning event in the fall, a heavily fouled collector may not function efficiently. This study indicates that fouling and predation to artificial spat collectors is minimal from June to August. Scallop recruitment during this time is not noticeably affected by the foul- ing of the spat collector. DESIGNING AN ACCESS SYSTEM FOR OCEAN MARI- CULTURE. Porter Hoagland and Di Jin, Marine Policy Center. Woods Hole Oceanographic Institute. Woods Hole. MA 02543. Ocean mariculture operations have been proposed as alterna- tives to traditional commercial wild harvests in the U.S. exclusive economic zone (EEZ). Unlike marine fisheries, ocean mariculture operations are designed to constrain the stocks being raised to specific geographic areas using nets. pens, or other technologies. The site-specific nature of ocean mariculture operations requires "security of tenure" (limited property rights) to designated areas of ocean space, possibly including the underlying seabed and neritic and surface waters. Although the allocation of exclusive or proprietary rights to ocean space will be a contentious issue, with- out security of tenure, the potential exists for other uses of the ocean to impinge upon mariculture operations. Further, the avail- ability of investment capital for ocean mariculture operations is likely to be extremely limited in the absence of security of tenure. The United States has sovereign rights in its EEZ over the exploitation of commercial living resources. Historically, the United States has exercised those rights through policies designed to manage wild, open-access fish stocks. At present, there is no coordinated policy in the United States governing the use of the EEZ for ocean mariculture operations. U.S. policy is not fully developed with respect to the siting of such operations, and per- mitting IS likely to proceed on an inefficient, ad hoc basis. A systematic approach to the design of an access system for ocean mariculture operations should involve the following steps: (1) drawing lessons from the design of access systems for other public resources; (2) examining historical practice and current op- eration of access systems in other jurisdictions; (3) developing a description of the resource to be allocated (ocean space), its rel- evant attributes, and any potential economic side-effects (positive or negative) that are likely to occur; (4) identifying relevant social objectives; (5) developing an analytic framework within which to analyze tradeoffs among the relevant social objectives; (6) positing a relevant set of policy attributes that would enable the specified social objectives to be met, including property right transfers (par- tial or complete, permanent or temporary); revenue generation; performance requirements (time limits, fees); information man- agement; environmental protection; and fairness or equity consid- erations; among others. DEVELOPMENT OF A GROUNDFISH AQUACULTURE INDUSTRY IN NEW ENGLAND: REVIEW OF EFFORTS TO DATE. Hunt Howell.' Linda Kling," Terry Bradley.' and- Larry Buckley,' 'University of New Hampshire, Department of Zoology, Durham. NH 03824; "University of Maine, Orono, ME 04469; 'University of Rhode Island, Kingston, RI 02881. The decline of groundfish populations has caused a renewed interest in growing cod, haddock and flounder species for food and/or stock enhancement. A number of research projects, aimed at raising groundfish species on a commercial scale, are currently underway. This presentation will review the research being done for cod and haddock. Summaries of studies done over the last year, as well as work planned for the future, will be included. HATCHERY CULTURE TECHNIQUES FOR THE GIANT SEA SCALLOP PLACOPECTEN MAGELLANICUS. Richard C. Karney, Martha's Vineyard Shellfish Group. Inc.. Oak Bluffs. MA 02557. Under funding from the National Marine Fisheries Service Fishing Industry Grants Program, the Martha's Vineyard Shellfish Group adapted hatchery culture methods for bay scallops to the successful culture of the giant sea scallop (Placopecten magellan- icus). Field-collected broodstock were sufficiently ripe in early March (sea water temperature, 4-5°C) to spawn just over seven million eggs. The fertilized eggs were transferred to a 400 liter larval conical, with l-|ji, filtered, aerated, seawater at 12°C. After 48 hrs. the conical was drained and about three million scallops (ciliated blastulae and trochophores) were recovered and resus- pended. Straight-hinge larvae were not observed until the second drain down on day 4. Larval culture protocol throughout the lengthy 40 day larval period included a daily feeding of Isochrysis sp (T-ISO) and/or Chaetoceros neogracile, with a drain down and sizing every other day. The larvae were cultured in three 400-liter conicals of 5 p. bag-filtered seawater, heated to about 15° C (range I3-I7°C). Between days 28 and 38, 1 ,350.000 pediveliger larvae (about 250 |jl) were moved to downweller sieves. The first fully set juvenile was observed on day 32. Set scallops were cultured on 456 Abstracts, February 26-28. 1996 Milford Aquaculture Seminar. Milford. Connecticut downweller sieves (130-300 \x.) with a flow of bag-filtered water (10-50 (x) at seawater temperatures of 8-16°C. By June 8. (day 90) the largest seed measured 2-nim and were moved to 1-mm mesh Korean spat bags in a cage anchored off the dock of the shellfish hatchery. By July 3. a total of 519,000 2-mm seed were successfully transferred to the inshore field culture systems. All of several potential off-shore growers were frustrated in their attempts to secure proper permits from the regulatory agen- cies. Over 80% of the 519,000 seed scallops were lost during the month of July when water temperatures reached 22°C and regu- latory delays prevented transfer of the seed to cold water grow-out sites. At the end of July, the remaining 90.000 seed were finally permitted to be moved to the deep water (65') sites of the Truro Aquaculture Project in Cape Cod Bay. OVERVIEW OF SEED HARD CLAM WINTER MORTAL- ITY STUDIES IN NEW JERSEY. NEW YORK AND MAS- SACHUSETTS. John Kraeuter.' John Aldred." Paul Bagnall.' Richard Crema/ Gef Flimlin."' Susan Ford,' Dale Leavitt,' George Mathls,'' Gregg Rivara.^ Roxanna Smolowitz,^ and Margy Winlermyer,' 'Haskin Shellfish Re- search Laboratory. Rutgers University. Port Norris. NJ 08349; ^Town of East Hampton. NY 1 1937; ""Town of Edgartown. Edgar- town. MA 02539; ^R. F. Crema and Family. Oceanville. NJ 08231; ^Woods Hole Geeanographic Institution. Woods Hole. MA 02543; "Mathis and Mathis Enterprises. Egg Harbor. NJ 08251; ''Cornell University Cooperative Extension. Riverhead. NY 11901; ^University of Pennsylvania and Marine Biological Laboratory, Woods Hole. MA 02543; '^New Jersey Cooperative Extension, Toms River. NJ 08753. Hard clam [Mercenaria mercenaria) aquaculture is one of the most widespread forms of marine aquaculture on the U.S. east coast. The industry is based on hatchery seed production and planting of these seed in protected beds for growout. In the North- east and mid- Atlantic, seed planted late in the season is subject to variable (up to 50% in some beds) and unpredictable mortalities. Similar observations have been made for manila clams on the Pacific coast. The reasons for the variable success of late plantings of 8-15 mm seed have not been systematically investigated, but are believed to be related to seed ""condition" (stored energy reserves) and its interaction with environmental variables or patho- gens. One solution to late planting mortality is to hold the clams in nursery systems during winter. Unfortunately, a second, but ap- parently related, mortality occurs when seed clams are overwin- tered in nursery systems. Once seed begin to die in these systems. losses can be as high as 5% per day. The current work is developing and testing methods that could be used by culturists to evaluate the "condition" of seed clams prior to planting or overwintering. The goal is to provide a quan- tifiable means of evaluating alternatives to mitigate losses and to integrate the methodology into an economic/biological decision matrix. The work underway will evaluate whether disease (bacte- rial or other infections), lack of energy reserves, or their interac- tion is the primary cause of winter mortalities. The overall exper- imental protocol is; 1 . Evaluate methods for assessing condition on specific size classes of seed. 2. Test seed and experimentally ma- nipulate condition of specific seed sizes. 3. Re-analyze these ma- nipulated seed and plant at Vi commercial scale. 4. Place addi- tional seed, at specific temperatures, during the fall in field nurs- ery systems. 5. Test growth, survival and condition of all seed in spring. This is New Jersey Experimental Station Publication No. K-32403-1-96 and a Contribution of the Institute of Marine and Coastal Sciences, Rutgers University. Supported by state funds and grants from the Northeast Regional Aquaculture Center. Na- tional Science Foundation, and the Fisheries and Aquaculture Technology Extension Center of Rutgers University. THE USE OF VACCINES IN THE CULTURE OF PENAEID PRAWNS. J. W. Latchford,' S. B. Prayitno,' and A. Alabi,' 'School of Ocean Sciences, University of Wales Bangor, Menai Bridge, Gwynedd, UK; "Department of Fisheries, Diponegoro University, Semarang 50242, Indonesia. The increased incidence of diseases in cultured shrimp cou- pled with a growing awareness of the problems of the use of antibiotics in controlling such diseases has led to the development of alternative methods of disease control. Vaccines against several strains of luminous and non-luminous bacterial pathogens were tested for their efficacy in both small scale and commercial scale culture systems of Penaeus iiuiicus and P. monodon larvae and postlarvae. Formalin-killed bacteria and vaccines consisting of live atten- uated strains of pathogenic bacteria produced by UV light mu- tagenesis gave a significant degree (p < 0.05) of protection against subsequent infection by virulent pathogens when compared to non- vaccinated controls in small scale culture systems. EXPERIMENTAL EVALUATION OF VIBRIO SPP. AS ETI- OLOGICAL AGENT OF JOD. Mijin Lee.' Gordon T. Tay- lor.' Monica Bricelj.' and Susan E. Ford.^ 'Marine Sciences Research Center. State University of New York at Stony Brook. Stony Brook. NY 1 1794; "Haskin Shellfish Research Laboratory, Institute of Marine and Coastal Sciences, Rutgers University, Port Norris, NJ 08349. The potential role of Vibrio spp. in juvenile oyster disease (JOD) was experimentally evaluated during the summers of 1994 and 1995 with isolates collected from juvenile oysters at the Frank M. Flower and Sons Oyster Co. during the JOD episode. To establish which of the isolated Vibrio induced JOD symptoms, six challenge experiments were performed at the Flax Pond Marine Laboratory of SUNY Stony Brook. For the first, second, and third challenge experiments, bacterial suspensions were delivered via injection into the mantle cavity of juvenile oysters (size of oysters = 17.5-35.4 mm). For the fourth, fifth, and sixth challenge ex- periments, oysters were exposed to high concentrations of bacte- Milford Aquaculture Seminar, Milford, Connecticut Abstracts. February 26-28. 1996 457 rial suspensions in tanl; water instead of being injected (size of oysters = 12.5-20.0 mm). All bacterial isolates caused mortality in excess of controls, but one isolate, phenotypically similar to Vibrio angiiillarum. consistently produced higher mortalities, within 7-14 days of exposure, than all other isolates. Conchiolin deposits similar to early JOD shell deposits were found on oysters in some experiments; however, they were not restricted to a single bacterial isolate. The results of these experiments clearly show that Vibrio spp. isolated from JOD-affected oysters, especially a strain similar to a V. anguilkintm. can be pathogenic to juvenile oysters under certain conditions, and that these isolates can elicit conchi- olin layering similar to JOD deposits. No single isolate produced clear or consistent JOD symptoms, however, suggesting that out- breaks of the disease may not be associated with a single bacterial strain but have a multiple-factor etiology. 1995 JUVENILE OYSTER DISEASE (JOD) TRANS- MISSION AND BACTERIOLOGICAL STUDIES. Earl J. Lewis,' C. Austin Farley," Ana Baya," and Rosan- gela Navarro,^ 'USDOC, NCAA. National Marine Fisheries Service, Cooperative Oxford Laboratory, 904 S. Morris St., Ox- ford, MD 21654-9724; "Maryland Department of Agriculture, Animal Health Diagnostic Laboratory, 8077 Greenmead Drive, College Park, MD 20740. Three laboratory. transmission experiments for JOD were con- ducted in 1995. As in previous experiments, material from 3000- 35(X) gallons of ambient water at a JOD-affected facility was fil- tered sequentially through bag filters (1-100 (im). Material was backflushed from each filter into 1 gallon of synthetic seawater and added to an aquarium containing susceptible oysters to test the relative infectivity of the material. In addition, 97 water samples and 280 oysters from field and experimental studies were cultured for bacteria. Samples for the first transmission experiment were collected in May 1995, 10-11 weeks prior to the onset of JOD. Water tem- perature at the time was 15°C. JOD was minimally transmitted by the 10 jjim material at room temperature (24-26°C). Mortalities in these oysters were only 4%, but 75% of the dead oysters had typical JOD conchiolinous shell lesions. Mean vibrio counts in aquaria water at the beginning of the experiment ranged from 10,000 to 1,000,000 colony forming units (CFUs)/ml. Unlike vibrios, the JOD infective agent apparently was not present in the water column in large quantities at this time of the year. From water sampled in July 1995, JOD was transmitted at room temperature (21-23°C) using material held in 5, 10, 25, 50 and 100 ^.m bag filters. Oysters held in the 10 and 25 \i,m filtered material experienced the highest JOD-related mortalities, 33 and 36%, respectively. Mean vibrio counts in aquarium water ranged from 23,000 to 160,000 CFUs/ml at the beginning of the experi- ment. An October 1995 experiment also transmitted JOD at room temperature, but the resulting mortalities were influenced by tem- perature fluctuations. Onset of JOD-related mortalities began in the 5, 10, 25, 50 and 100 |jim exposures after 2-3 weeks, but did not progress after the water temperature dropped to 12°C. Mor- talities and conchiolinous shell lesions began to reappear 2-3 weeks after heaters were placed in aquaria to maintain tempera- tures above 20°C. Mean vibrio counts ranged from 4000 to 28,000 CFUs/ml. Disease transmission was not evident in oysters exposed to 1 or < 1 fjLm material from any of the experiments. Bacteriological cultures of oysters and water samples from lab- oratory transmission experiments and regional samples from New York, Maine, and Rhode Island yielded 17 known Vibrio spp. As in our previous work, there was no consistent association of a particular Vibrio sp., or group of vibrios, with JOD. All com- monly isolated vibrios (prevalence >9%) were isolated from un- infected control samples and JOD-infected samples. Vibrio sp. were isolated with neariy the same prevalence from each group. DIAGNOSIS AND PREVALENCE OF PERKINSUS SP. IN CHESAPEAKE BAY SOFTSHELL CLAMS, MYA ARE- NARIA: AN UPDATE. Shawn M. McLaughlin, USDOC, NOAA, National Marine Fisheries Service, Cooperative Oxford Laboratory 904 S. Morris St., Oxford, MD 21654-9724. Perkinsus sp. was rarely observed in Chesapeake Bay soflshell clams (Mya arenaria) prior to 1990 based on histological analyses. An unusual occurrence of the parasite was observed in Chesapeake Bay softshell clams from seven sites at prevalences ranging from 3-53% during 1991-1993 based on rectal assays in fluid thiogly- colate medium (FTM). Prevalences decreased in 1994 and re- mained low at sites examined in February 1995. In August 1995. softshell clams from three Chesapeake Bay sites were diagnosed by rectal FTM assays to have significant prevalences of Perkinsus sp. Prevalences of the parasite were 13% at Swan Point, 37% at Piney Point, and 64% at Cedar Point. Analyses of tissue sections indicate the initial infection site of softshell clam Perkinsus sp. is often the gills. A comparison of softshell clam tissues incubated in FTM was conducted. Subsamples of hemolymph, gill demi- branchs, and rectal tissue collected from each of 30 Swan Point softshell clams were incubated separately in FTM for 5 days, stained with Lugol's iodine, and examined. Results showed Per- kinsus sp. prevalences of 0% for hemolymph. 13% for rectal tis- sue, and 90% for gill tissue in the Swan Point sample. The ex- periment was repeated using hemolymph. gill, rectum, and labial palps collected from 30 additional Chesapeake Bay softshell clams. Prevalences of Perkinsus sp. were 0% for hemolymph. 43% for rectal tissue. 90% for gill tissue, and 100% for labial palps. Softshell clam labial palps are now routinely incubated in FTM during histologic processing at this facility. The change from rectal to palps thioglycolate tests increases the sensitivity of the softshell clam Perkinsus sp. assay and eliminates the time con- suming task of locating and excising softshell clam rectal tissue. This study also indicates under-reporting ot Perkinsus sp. in Ches- apeake Bay softshell clams has occurred. 458 Abstracts. February 26-28. 1996 Milford Aquaculture Seminar, Milford. Connecticut FIG'S AND AQUACULTURE— A LOOK BACK AND A LOOK AHEAD. Harold C. Mears, USDOC. NCAA, National Marine Fisheries Service, One Blackburn Dr. Gloucester, MA 01915. In March 1994, the Secretary of Commerce announced the availability of $30 million under provisions of the Northeast Fish- eries Assistance Program to address the needs of those directly affected by the decline of the traditional fisheries in the Northeast. The initiative included $9 million for Fishing Industry Grants (FIGs) administered by the National Marine Fisheries Service. Aquaculture-related investigations comprise $4. 1 million, or about 46%, of the total funding under the FIG Program. The project objectives for these 22 studies are to help restore overfished groundfish and shellfish stocks in the Northeast Region, as well as to provide new business opportunities for displaced fishermen. Study activities are focusing on cod, haddock, summer floun- der, Nori (Porphyra). scallops, quahogs, oysters, surf clams, and sea urchins. The work funded under this program is providing an impetus for facilitating resolution of issues concerning aquaculture start-up costs, technology transfer, environmental impacts, user group conflicts, fishery enhancement, and potential for industry expansion and employment. The knowledge gained from the FIG Program is forging more effective communications and partner- ships between the private, state, federal, and academic sectors, as well as further specifying research and management priorities for future projects concerning aquaculture development. CULTURE OF THE BAY SCALLOP, ARGOPECTES IRRA- DIANS, WITHIN A SMALL-BOAT MARINA ON LONG IS- LAND SOUND (CONNECTICUT). Matthew Mroczka,' Paul Dinwoodie,' Ronald Goldberg." Jose Pereira," Paul Clark,* Sheila Stiles," Joseph Choromanski." Daniel Schweitzer,""' and Nancy Balcom,"* 'Cedar Island Marina, P.O. Box 181, Clinton CT 06413; "USDOC, NOAA, National Manne Fisheries Service, Northeast Fisheries Science Center, Milford Laboratory, Milford CT 06460: 'Marine Sciences and Technology Center, University of Connecticut, Groton, CT 06340; "^University of Connecticut, Sea Grant, Marine Advisory Office, University of Connecticut, Groton, CT 06340. An innovative suspension-culture rack system was designed to evaluate the potential of intermediate grow-out of shellfish seed withm a marina in Clinton. Connecticut. Conventional dock space in the marina was modified by cutting out sections of the decking to gain access to the water below. The cut-out sections were re- placeable, allowing normal use of the dock. Wire-mesh cages ( 1 x 0.5 X 0.5 m) containing four shelves were suspended below the modified docks. Shellfish seed were contained on the shelves of the cages within flexible plastic-mesh bags with temporary clo- sures of slit PVC pipe at the ends. To evaluate the growth potential if hatchery-reared bay scallop seed in the marina environment, aliout 6,000 animals with an initial shell height of 15.5 mm were reared at different densities and in different mesh size cages and bags from June through November of 1995. Survival of scallops in all treatments was very high, averaging about 90 percent. The fastest growing group of scallops reached an average shell height of 54. 1 mm by the end of November, with many individuals larger than 60 mm. Stocking density and mesh size were inversely pro- portional (P < 0.05) to growth of scallops over a wide range of sizes. Seawater current tlow, temperature and oxygen regimes, and ambient phytoplankton densities were adequate at this location to support substantial growth with low mortality. The use of space under the docks of marinas for shellfish culture caused limited interference with marina activity. There is a good potential for intermediate grow-out of scallop seed at marinas as a step in in- tensive aquacultural production or in seed transplant efforts to restore scallop fisheries to natural habitats. This work was partially funded through grants from the Uni- versity of Connecticut's Marine Technology Center and Sea Grant, Marine Advisory Program, Groton, CT. CULTURE OF SUMMER FLOUNDER, PARALICHTHYS DENTATVS, AT GREATBAY AQUAFARMS. George Nardi, GreatBay Aquafarms, 153 Gosling Road, Portsmouth, NH 03824. GreatBay Aquafarms, Inc. (GBA) was established in 1995 for the culture of marine fish, principally summer flounder. The initial objective of GBA is the production of 5-10 gram juveniles from its Portsmouth, NH hatchery. The hatchery is a 10,000 square foot facility which had been a warehouse for the Public Service Com- pany of New Hampshire, an electric utility. The hatchery began production in January of 1996. The facility includes a lab, phy- toplankton and zooplankton culture rooms, and three recirculating systems for the broodstock, weaning and nursery stages. The egg and early larval stages are flow-through water systems. At capac- ity, the hatchery is expected to produce between 300,000 and 400,000 juveniles per year. Including office staff, seven people are employed at GBA. Our production techniques are based both on university re- search undertaken here in New England and on a transfer of tech- nology from the European flatfish culture industry. GBA main- tains 5 stocks of broodfish which will be induced to spawn through photoperiod and temperature manipulation. In addition to the cul- ture of rotifers and artemia, GBA will initiate the culture of cope- pods as an early larval diet. The hatchery employs a sophisticated direct digital control system produced by Allerton Technologies. GBA will provide growers with juveniles and will work with them to assist in the development of this new industry. GBA also plans to establish its own grow-out farm in the near future. BAY SCALLOP CULTURE IN A VIRGINIA SALTWATER POND. Michael J. Oesterling and Laura A. Rose, Virginia In- stitute of Marine Science, College of William and Mary, Depart- ment of Advisory Services, Gloucester Point, VA 23062. Efforts at bay scallop culture (Argopecten irradians) in Vir- ginia begun in the 1960's by Mike Castagna, experienced a resur- gence in 1990. At that time, lantern net technology for grow-out Milford Aquacullure Seminar. Milford. Connecticut Abstracts. February 26-28. 19% 459 was not considered feasible, primarily due to regulatory uncer- tainty surrounding the permitting of large numbers of lantern nets in Chesapeake Bay. As the result of increasing publicity, the owner of a saltwater pond requested the evaluation of his pond for scallop culture. Private ownership of the pond made the use of hanging culture possible without any regulatory constraints. Periods of growth over the duration of this project can be divided into three phases. Anmials with a mean shell height of 4.7 mm were stocked into I -mm mesh pearl nets on 22 June 1995 and cultured at a very high density for 21 days (Phase 1). They were subsequently culled, restocked and grown at a density of 4388- 5265 per square meter (406-488/sq ft) for 38 days (Phase 11). For final grow-out. they were transferred to either 6-mm mesh pearl nets (439 sq m or41/sq ft) or 15-mm mesh lantern nets (Phase 111). Lantern nets were either 4-tiered or 5-tiered, with each tier holding 100 animals (density of 5 1 3 sq/m or 50 sq ft) and were stocked on 25 August and 1 September. Pearl net growth during Phase 1 averaged 0.20 tnm per day. Growth in pearl nets at mid-density during Phase II averaged 0.32 mm per day. Final grow-out in the pearl nets at a reduced density averaged 0.33 mm per day for 81 days. Lantern-net growth over 1 10 days averaged 0.34 mm per day. However, during September and early October, growth in lantern nets averaged 0.65 mm per day. In both pearl nets and lantern nets, market-size animals (over 40.0 mm shell height) were produced by the first week of October, approximately 100 days after initial stocking. tory for highly pathogenic larval pathogens, but inadequate when studying slower acting or more opportunistic bacterial infections. Large numbers of test animals were kept alive for up to three months in an inexpensive, easily maintained system using aerated 3-liter plastic jars. In addition, procedures were devised to facil- itate feeding, water changes, and observation while at the same time minimizing bacterial contamination of the surroundings. This arrangement would also be suitable for maintaining other species of bivalves and allow other types of studies such as juvenile oyster disease, toxicology, and viral transmission. This project was funded partially by a University of Connect- icut. Marine Sciences and Technology Center Grant. LESSONS FROM SCALLOP SPAT COLLECTION IN WASHINGTON STATE. Walter Y. Rhee, School of Fisheries, University of Washington WH-10, Seattle. WA 98195. Japanese onion-bag collectors were set in Hood Canal. Wash- ington, to test the feasibility of scallop grow-out culture through natural spat collection. A total of 406 fine-mesh (2 x 1 mm weave) onion bags were deployed for a period of three months from April to June, each onion bag holding one of three types of substrates: netlon. discarded gill nets, or acetate film. The results and the lessons learned from the collection are discussed. LABORATORY CULTURE OF TAUTOG: A PILOT STUDY. Dean M. Perry, Renee Mercaldo-AIIen, Cather- ine Kuropat, and James Hughes, USDOC. NCAA. National Marine Fisheries Service. Milford Laboratory. Milford. CT 06460. Spawning of field-captured adult tautog (Tautoga onitis) was accomplished in the laboratory. The embryos were cultured to hatching and successfully raised through the difficult larval stages to juveniles. Static culture containers, changed twice a week, proved superior to flow-through seawater systems. Newly hatched larvae were fed protozoans for the first 4 days post-hatch, and then were fed rotifers and anemia that had been enriched with highly unsaturated fatty acids. Larval mortality was high until natural plankton was added to the diet. Laboratory cultured artemia. sup- plemented with natural plankton and a commercial food made an adequate diet for juvenile tautog. A SIMPLE SYSTEM FOR LONG-TERM EXPOSURES OF ADULT OR LARVAL BIVALVES TO BACTERIAL PATHOGENS. Steven Pitchford, USDOC. NCAA. National Marine Fisheries Service. Northeast Fisheries Science Center. Milford. CT 06460. A method used for long-term exposures of larvae, juveniles and adult bay scallops (Argopecten irnuUans) to bacterial pathogens is described. The standard 48-hour static screening tests are satisfac- CHANGING TIMES FOR RHODE ISLAND SHELLFISH- ERIES: URI's SHELLFISH AQUACULTURE TRAINING PROGRAM FOR SHELLFISHERMEN. Michael A. Rice and Joseph T. DeAlteris. Department of Fisheries. Animal and Vet- erinary Science. University of Rhode Island. Kingston. Rl 02881. The shellfishing industry in Rhode Island has been in decline since the middle 1980s. In 1985, about 27 million pounds (shell on) of northern quahogs, Mercenaria mercenaria, were harvested from Rhode Island waters, but by 1993 only 11 million pounds were harvested. In the same time period, the number of full-time shellfishermen dropped from about 800 to a present number of 300. Catch per unit effort by remaining fishermen is about the same as in the previous decade, but market prices have remained low leading to depressed fisheries incomes. Traditionally, the Rhode Island shellfishing community has been reluctant to em- brace aquaculture activities in the state, but recent political action by the State Legislature has brought aquaculture discussion to the forefront. A legislative commission charged with the promotion and protection of an aquaculture industry in Rhode Island has been formed, and aquaculture-friendly legislation is pending. Although opposition to aquaculture remains among some shellfishermen. other shellfishermen are considering the adoption of shellfish 460 Abstracts . February 26-28, 1996 Milford Aquaculturc Seminar, Milford, Connecticut aquaculture as a supplemental income source. In the summer of 1995, we conducted a course for shellfishermen using the tech- nique of transient gear aquaculture of oysters, Crassostrea virgin- ica. A total of 17 students registered for the course, 12 of whom were shellfishermen. Lectures covered issues of permitting, oyster biology, business planning, and marketing. Using the URl vessel R/V Captain Bert, the shellfishermen deployed cages at three sites in Narragansett Bay on June 7, using about 150,000 3^ mm seed (=750 ml volume) oysters per site. Monitormg the three sites (Dutch Island Harbor, Hope Island and Fox Island) on a biweekly basis, the students found that oysters grew at varying rates de- pending on site. At the Fox Island site, a total volume of 217 1 of oysters was attained by October 30 with sizes averaging 30-40 nmi valve height. At Dutch Island Harbor and Hope Island oyster biovolumes were 69 1 and 123 1, respectively, with minimal mor- talities. The growth of oysters at Fox Island approaches that of oysters in the more nearly eutrophic Point Judith Pond. Although fouling of enclosures is less in the Narragansett Bay sites than in the pond, starfish predation is greater in the Bay. These results were pumped through a 60-(jLm plankton net and preserved in cally feasible in Narragansett Bay, but attention must be paid to effective predator exclusion. This is publication number 3210 of the College of Resource Department, University of Rhode Island. MARTHA'S VINEYARD SHELLFISH GROUP AQUACUL- TURE TRAINING PROGRAM. Elizabeth F. Scotten, Gabri- ella C. Castro, and Debra L. Colombo, Martha's Vineyard Shellfish Group, Inc., Oak Bluffs, MA 02557. Under funding from the National Marine Fishenes Service (NMFS), Fishing Industry Grants (FIG) Program and a NMFS grant to the Nantucket Research and Education Foundation (NREF), the Martha's Vineyard Shellfish Group initiated an Aquaculture Training Program for fishermen who have been dis- placed by the George's Bank fishing area closures. The program was seen as an opportunity for fishermen to make a transition from offshore fishing to shellfish aquaculture business. A total of 15 trainees worked closely with the staff of the Martha's Vineyard Shellfish Group and the local shellfish consta- bles learning hatchery and on- and off-shore nursery techniques. The trainees participated in all aspects of larval and juvenile care, from algal culture and spawning to grow-out, both on- and off- shore. The trainees assisted with the design, assembly and oper- ation of an on-shore shellfish culture nursery. They also spent time with the local shellfish constables building off-shore rafts and monitoring growth therein. The program also included a weekly lecture series where sci- ence, industry and regulatory people addressed many of the issues facing aquaculturists today. By providing exposure to numerous aquaculture resources and through a hands-on approach to aqua- culture, the Martha's Vineyard Shellfish Group developed a po- tentially effective aquaculture training program. A PC-CONTROLLED, EXPERIMENTAL MOLLUSCAN REARING APPARATUS FOR STUDIES OF FEEDING STRATEGIES AND NUTRITION. Barry C. Smith,' Gary H. Wikfors,' Jennifer H. Alix," and Mark S. Dixon, '■'^ 'USDOC, NCAA, Northeast Fishenes Science Center, Milford Laboratory, Milford, CT 06460; "University of Connecticut, Ma- rine Sciences and Technology Center, Groton. CT 06430. Production of microalgal feeds has been a major limiting factor in the development of the shellfish aquaculture industry. On a production scale, many grams per week of microalgal biomass may be required to feed shellfish. Knowledge of feeding regimes yielding the highest conversion efficiencies of algal feed to mol- luscan growth is required to maximize the return on an algal- culture investment. At the Milford Laboratory, specialized, manually-controlled molluscan rearing chambers have been used to study shellfish nutritional requirements since 1982. The system consists of twelve PVC chambers fitted with screens to hold the shellfish being stud- ied. Contemporary process-control technology now can be used to program such things as feeding time and duration in investigations of shellfish nutrition. A computer-controlled, solenoid-valve sys- tem has been added to the existing manual system to control sea- water flow, volume of microalgal food, and feeding duration au- tomatically. All components of the system are "off-the-shelf" in that they are readily available. An object-oriented software pack- age controls the outputs of a digital In-Out board; the output sig- nals trigger relays which operate the solenoid valves. Each cham- ber has a solenoid valve for seawater flow, algal feeding, and chamber draining. The system runs independently until stopped, reducing labor requirements while adding experimental flexibility. Each chamber represents a model for a programmed molluscan nursery system. Initial results suggest that feeding young post-set bay scallops, Argopecten irradians. every six hours yields growth superior to feeding once daily. Subsequent experiments will work toward de- veloping feeding standards for molluscan shellfish analogous to those employed routinely in agricultural animal husbandry. This project was funded partially by a University of Connect- icut, Marine Sciences and Technology Center Grant. AN IMPORTANT NEW DISEASE OF HARD CLAMS, MERCENARIA MERCENARIA, IN THE NORTHEAST UNITED STATES. Roxanna Smolowitz,' Dale Leavitt,' and Frank Perkins,"' 'Laboratory for Marine Animal Health, Univer- sity of Pennsylvania, Marine Biological Laboratory, Woods Hole, MA 02543; ^Department of Biology. Woods Hole Oceanographic Institution, Woods Hole, MA 02543; 'Department of Zoology, North Carolina State University, Raleigh, NC 27606. In July of 1995, a meeting of hard clam culturists was held with Drs. Leavitt and Smolowitz in Provincetown, MA. The culturists described a 4-year history of chronic, increasing severe clam mor- Milford Aquaculture Seminar. Milford. Connecticut Abstracts. February 26-28, 1996 461 tality on the planted flats in Provincetown. MA. In 1995. up to 80% of the seed planted 1-1/2 to 2 years before was dead. Crab predation was a recognized problem on the clam leases, but the additional possibility of a primary disease was considered since losses appeared to be too high for crab predation alone. Clam samples were obtained at that meeting and at a subsequent visit by Smolowitz and Leavitt to the leases in August. Histologic exam- ination of 12 clams showed 6/12 contained an endosporulating protozoal organism similar to QPX (seen in a Canadian hatchery in 1989). These Provincetown animals, however, also showed bac- terial infections. Leavitt and Smolowitz revisited two Provincetown leases in mid-October. 1995. A total of 80 clams were collected. Fifty clams were identified as moribund or " "affected" (poor growth over the summer, slight gaping of 1-2 mm and chips in the shell edges). Thirty animals from the same flats were collected as con- trols (at least 1 cm of new growth over the summer and no gaping or chips). Histopathologic examination showed that 45/50 of the "affected" clams contained the QPX-like parasite accompanied by severe inflammation. The parasite was seen in only 3/30 of the control clams. No other possible disease-causing agent was seen in these clams. In November 1995. both Smolowitz and Perkins were sent samples of approximately 18-month-old clams from a culture site in Duxbury. MA. .One population of clams from that site had experienced heavy mortalities over the preceding 3 months. Ex- amination of these clams showed that the same QPX-like organism was responsible for the mortalities in Duxbury. MA. This Work is a result of research sponsored by NOAA National Sea Grant College Program Office. Department of Commerce, under Grant No. NA46RG0470. Woods Hole Oceanographic In- stitution Sea Grant project No. R/A-33-FO and funding from the WHOI Coastal Research Center Rapid Response Program. size of scallops in the high or fast growth line larger than that of scallops in the slower-growing line. Scallops were also sampled for shell color such as stripes or or- ange, and were self-fertilized to determine inbreeding effects, as well as to explore the potential for improvmg heritable production traits that might be linked to shell markers. Markings were inherited. How- ever, in contrast to mass-spawned cultures, development to 48 hours and subsequent survival were lower and growth retarded in several inbred scallop cultures, suggesting inbreeding depression. Breeding results were supported by a population genetics in- vestigation of wild and cultured bay scallops employing enzyme electrophoresis conducted cooperatively under the sponsorship of the U.S. -China Marine and Fisheries Protocol Agreement on Liv- ing Marine Resources. Scallops obtained from various populations ranging from Canada to Florida, as well as from Milford Labora- tory strains, were measured and characterized for size, age, rib number, and shell and mantle color. Specimens were then sampled for enzyme electrophoresis and eventual DNA analysis. Other samples were brought from China or purchased from a supermar- ket. Allozyme analysis revealed genetic variation among and within populations, which should be considered in broodstock se- lection and enhancement and restoration efforts for aquatic organ- isms in general, including finfish. Milford mass-spawned and in- bred lines provided valuable information regarding genetic diver- sity and stock structure. Field assessments of habitat suitability and strain performance in stock restoration and enhancement efforts are indicating growth of some scallops to marketable size in less than a year. All of these projects demonstrate that the bay scallop is a suitable model for genetic studies and that it can have excellent responses to selection for improved growth. This project was funded partially by a University of Connect- icut. Marine Sciences and Technology Center Grant. PRELIMINARY INVESTIGATIONS OF GENETICS AND BREEDING OF THE BAY SCALLOP, ARGOPECTEN IR- RADIANS. Sheila Stiles,' Joseph Choromanski,' Daniel Schweitzer.'- and Qin-Zhao Xue,' 'USDOC. NOAA. National Marine Fisheries Service. Milford Laboratory. Milford. CT 06460; "Marine Sciences and Technology Center. University of Connecticut. Groton. CT 06340; 'Chinese Academy of Sciences, Qingdao. China. Genetic investigations that comprise various approaches for increasing growth rate of the commercially valuable bay scallop, Argopecten irradiaiis, were initiated. Projects included mass- selection, inbreeding, population genetics and strain/field evalua- tions. Approximately 15 mass spawnings were accomplished to establish foundation crosses for mass-selection. Scallops were measured and the top 10-15% or largest scallops and the bottom 10-15% or smallest scallops were selected and then spawned to produce the next generation. Preliminary results from initial se- lective breeding for fast growth are indicating success with mean VELIGER VODOO AND OTHER WITCHCRAFT IN THE WESTPORT RIVER: BAY SCALLOP, ARGOPECTEN IR- RADIANS. VELIGER ABUNDANCE AND THE VARIABIL- ITY OF SPATFALL RECRUITMENT TO ARTIFICIAL SPAT COLLECTORS IN THE WESTPORT ESTUARY, MASSACHUSETTS. Karin A. Tammi,' Wayne H. Turner,^ Margaret Brumsted,"' and Michael A. Rice,' 'Department of Fisheries, Animal and Veterinary Science, University of Rhode Island, Kingston, Rl 02881: "Water Works Group, P.O. Box 197, Westport Point, MA 02791: 'Dartmouth High School, Slocum Road. Dartmouth. MA 02747. During the past 3 years, the Bay Scallop Restoration Project has pioneered the widespread use of artificial spat collectors (4- mm plastic mesh onion bags containing monofilament) in the Westport River. Researchers have observed a significant variabil- ity in scallop recruitment to collectors deployed throughout the estuary. In general, poor scallop recruitment to spat collectors may be attributed to estuarine circulation patterns, crab predation and 462 Abstracts. February 26-28. 1996 Milford Aquaculture Seminar, Milford. Connecticut fouling. This information lias led to an investigation of the bivalve larval dynamics in the Westport River with the goal of optimizmg deployment time of spat collectors. From May to September 1995, weekly bivalve larval sampling was conducted within four main channels of the Westport Estuary. At each site. 200 liters of water were pumped through a 60-|j,m plankton net and preserved in ETOH for later enumeration and identification. Beginning in June, multiple longlines containing 20 spat collectors were deployed weekly to 5 study areas of which 3 were being monitored simul- taneously for bivalve larvae. At Corey's Island, an additional col- lector with a commercial fine mesh (1 .5 to 3.0 mm) contammg a polyethylene tube (40 x 80 cm) as the settlement substrate was also deployed. During the summer of 1995. 4 major bivalve spawning events occurred coinciding with full and new moons. Larval bivalve identification determined that these spawning events were bay scallops, blue mussels and soft-shelled clams. Monthly scallop recruitment to collectors reflected the larval bi- valve profile of that study area. The greatest scallop recruitment for all study areas was observed in longlines deployed in late June and harvested by mid-August. Corey's Island displayed the great- est recruitment (per 60 bags), yielding a total of 887 scallops, followed by Canoe Rock (250 scallops) and Horseneck Channel (230 scallops). Although the Hicks Cove and Jug Rock study areas were not sampled for bivalve larvae, the monthly scallop recruit- ment occurred at the same time as the other 3 areas. Jug Rock yielded a total of 862 scallops and Hicks Cove 108 scallops for the same dates. Corey's Island had the greatest overall monthly re- cruitment for the entire study due to the significantly better per- formance of the fine-mesh collector. This study indicates that A. irradians spawns heavily in late June and that the optimal time to deploy spat collectors is in late June with retrieval by mid-August, thus improving collection efforts for the future. UPDATE ON FEDERAL POLICY AFFECTING MARINE AQUACULTURE IN THE EXCLUSIVE ECONOMIC ZONE OF THE UNITED STATES. Eric M. Thunberg, USDOC. NOAA. National Marine Fisheries Service. Northeast Fisheries Science Center. Woods Hole. MA 02543. Marine aquaculture in the United States has been primarily located in coastal near-shore waters. However, for a variety of reasons development of coastal sites continues to be difficult prompting some aquaculture professionals to consider the feasi- bility of off-shore aquaculture. Recently, the New England Fish- ery Management Council approved a proposal to allow develop- ment of an experimental off-shore sea scallop demonstration proj- ect. A proposed salmon operation was considered but not approved. Proposals for ocean ranching of tunas and other species are still in the developmental stages but may eventually become a reality. In most instances, accommodation of aquaculture in the Exclusive Economic Zone of the United States will require mod- ification of existing law and modification of existing fishery man- agement plans. At this time, considerable ambiguities exist con- cerning the treatment of marine aquaculture in the EEZ which have only recently begun to be taken up by the United States Congress and Fishery Management Councils. This paper provides an update and status report on developing legislation and policy statements that will affect marine aquaculture in waters under Federal juris- diction. Unresolved issues are highlighted with special emphasis on their economic implications. WHAT A YEAR TO BE A MUD CRAB! THREE YEARS ON THE BAY SCALLOP RESTORATION PROJECT, WEST- PORT RIVER ESTUARY, MASSACHUSETTS. Wayne H. Turner, Karin A. Tammi, Bart Harrison, and Bethany A. Starr. The Water Works Group. Inc. , Post Office Box 197, West- port Point. MA 02791. The Bay Scallop Restoration Project IBSRP) was launched in 1993 with the goal of generating the interest, involvement, and en- thusiasm required to restore and enhance the renewable economic resources of traditional fishing and farming communities. Eighty- three thousand hours of volunteer work enthusiastically invested by teams of people have been channeled into this effort. These people: students, teachers, parents, graduate students have left a major im- pact, not only on the bay scallop. Argopecten irradians. but on the way in winch communities participate and positively affect the di- rection of their economic future and environmental quality. With three years of research on the BSRP. community volun- teers, led by graduate students have uncovered several significant clues about bay scallop propagation in the Westport River. In 1993, when the BSRP first began, mud crabs, Panopeus spp., went largely unnoticed as very few were found on the river bottom or in the propagation equipment. Green crabs. Carcinus maenas, on the other hand, were plentiful and practically every spat bag (propagation equipment used to catch juvenile scallops) had at least one if not two green crabs associated with it. Strangely enough, in the summer of 1994, green crabs were reduced in number and mud crabs surged. Researchers began counting thousands of mud crabs pouring from nearly every spat bag. Because of the recent prevalence of mud crabs, an experiment using floating rafts was set up in the Westport River, with each raft housing a different combination of four ingredients: mud crabs, green crabs, spat-size bay scallops, and yearling tautog, Tautoga onitis (approximately two inches in length). The conclusions drawn from this study are intriguing: 1 ) mud crabs ate the bay scallops; 2) green crabs did not eat the bay scallops and instead ate the mud crabs: 3) yearling tautog could not seem to handle a green crab (probably due to size of the crab), but cleverly enough, researchers observed that mud crabs in a raft with two-inch tautog lost one leg per day. By the fourth day of the experiment, the mud crab could no longer move and the tautog ate it. Therefore. 1995 appeared to be a good year to be a mud crab for at least three reasons. First, green crabs have been down in num- bers for the past two years. Secondly, yearling tautog, commonly found in spat bags in 1993. were virtually absent during 1994 and Milford Aquaculturc Seminar, Milford, Connecticut Abstracts. February 26-28, 1996 463 1995. Finally, support of this abundant supply of mud crabs seemed to become the propagation activities of the BSRP by sup- plying spat-size bay scallops as a preferred food source of mud crabs. GROWTH OF BAY SCALLOPS, ARGOPECTEN IRRADl- ANS, IN 5 MM MESH LANTERN NETS. James C. Widman, Jr.' and Christopher G. Cooper,'-^ 'USDOC, NCAA, National Marine Fisheries Service, Northeast Fisheries Science Center. Mil- ford, CT 06460; "Sea Change Foundation, Covington, VA 24426. One growing scenario for bay scallop culturists is to spawn during the natural warm-weather season, thereby reducing hatchery energy costs. If net-suspension culture is used, one needs to determine when biofouling would require cleaning. It is also impwrtant to determine whether any differences in growth occur due to handling, or by po- sition of the scallops in the shelves of the nets. To answer these questions we conducted the following experiment: Bay scallops, Argopecten irradians irradians. were deployed at a density of 750/m" in 5 mm mesh, 5-tiered lantern nets in Long Island Sound near Groton CT. Scallops with an initial mean shell height of 17.6 mm (8.1-32.5 mm range) were held from July 1 1 through November 27, 1995. Fifteen nets were deployed in a modified latin square design to determine the effects of handling, shelf position, and biofouling on scallop growth. Three nets were sampled every month, and the remaining 12 were sampled three per month. Each time a net was sampled the scallops were moved into a new net. Final mean shell heights ranged from 39.0-47.4 mm. Although there was a significant difference in mean shell height among shelves, the difference was not consistent, i.e.. shelf one did not always provide optimal growth. Scallops reared in the bottom shelf had lower survival because the bottom of the net often hit the cement anchor due to the high current. Scallops handled on a monthly basis were usually smaller and had lower survival than those handled only once during the experiment. Very little fouling of the scallop shells was observed, although the nets were always fouled. There was no effect of fouling on growth during the study period. The scallops harvested from this experiment are now enjoying winter at the bottom of Long Island Sound in Groton and Milford, CT. We plan to report the overwintering and final grow-out results next year. This project was funded partially by a University of Connect- icut, Marine Sciences and Technology Center Grant. FEEDING STRATEGIES FOR POST-SET BAY SCAL- LOPS, ARGOPECTEN IRRADIANS: WHAT? HOW MUCH? HOW OFTEN? Gary H. Wikfors,' Barry C. Smith,' Jen- nifer H. Ahx.' and Mark S. Dixon.'" 'USDOC, NCAA, Na- tional Marine Fisheries Service, Northeast Fisheries Science Cen- ter, Milford, CT 06460; "Marine Sciences and Technology Center, University of Connecticut, Groton, CT 06340. Bringing bay scallops to market in one growing season in the northeast will improve the economics of scallop farming from two standpoints: 1) losses to winter-kill will be avoided, and 2) the farmer's return on investment will not be deferred. To accomplish this, seed scallops will need to be produced early in the season in a heated, recirculated seawater system; it will take one heck of a lot of algae to feed them. Even more algal feed will be needed to grow scallops to market in such a system. The economics of prod- uct value versus feed costs will be dependent upon two major factors: 1 ) cost of algal biomass, and 2) feed conversion efficiency (i.e., how much algal biomass is converted to scallop biomass and how much is wasted?). Maximizing feed conversion efficiency requires answers to the three questions posed in the title. We have conducted several feeding experiments with young, post-set scallops comparing algal diets and feeding schedules to begin finding answers. "What" to feed post-set scallops seems to include; I ) cells larger than about 6 ^.m, 2) algal cultures that are not "clumped" ■ in aggregates, and 3) algal strains with high levels of total lipid and essential fatty acids. High-lipid Tetraselmis strains that support rapid growth of oysters are also superior scal- lop diets. The "How much" question has yielded less clear an answer. Doubling a ration that is only partially consumed results in sig- nificantly faster growth. This feeding behavior, however, will re- sult in wasted algal feed if rations are adjusted to maximize growth. One solution for this problem may be to find alternative uses for unconsumed algae from scallop-rearing tanks. An experiment to determine "How often" to feed post-set scallops yielded better growth when animals were fed every six hours, as compared with feeding less often or feeding every three hours. Differences between good and poor algal diets were less severe when fed at the optimal six-hour regime than when fed only once every 24 hours. These preliminary findings present logical directions for future research to develop biochemically-based feeding standards and schedules for nursery culture and grow-out of bay scallops. This project was funded partially by a University of Connect- icut, Marine Sciences and Technology Center Grant. PIONEERING EFFORTS TO PRIVATELY CULTURE QUAHOGS, MERCENARIA MERCENARIA, IN THE TOWN OF EDGARTOWN, MA. Paul Willoughby and Jack Blake, Martha's Vineyard Shellfish Group, Inc., Oak Bluffs, MA 02557. Fishermen participating in the Martha's Vineyard Private Aquaculturc Initiative, an aquaculturc training program funded un- der the Fishing Industry Grants Program of the National Marine Fisheries Service (NMFS) and a NMFS grant to the Nantucket Research and Education Foundation, describe their first attempts to field-culture quahogs in private and cooperative public/private projects in the town of Edgartown. One-half million small (1.5-mm) seed quahogs purchased on June 13 were successfully cultured to 20-mm in experimental nurs- 464 Ahslracts. February 26-28, 1996 Milford Aquaculture Seminar. Milford. Connecticut ery floats on existing and proposed private aquaculture lease sites. One raft design that performed well in an exposed site subject to 50 mph winds and waves up to two feet is described. Quahogs (6-10 mm) culled from the rafts at the end of July and planted in a muddy substrate were observed after two weeks to have suffered high (95%) predatory mortality by mud crabs. Larger (19 mm) seed planted in sandy bottom in September showed next to no mortality when sampled two weeks later. In mid-October 280.000 20-mm quahogs from the public/private project were seeded by ten fishermen at various Edgartown sites. Similar growth and survival was achieved in floating nursery trays on a private lease site where 160.000 of an initial 20O.(X)0 1 .5-mm seed quahogs measured 20 mm at the end of November. Difficulties experienced in securing private aquaculture lease areas are also presented. MURPHY'S LAW AND THE RAISING OF ATLANTIC STURGEON. Frederick B. Wishner, Hofstra University, Hempstead, NY 11551. The U.S. Fish and Wildlife Service Hatchery at Lamar. PA was successful in hatching 150.000 Atlantic sturgeon larvae on 4 July 1995 using a large female which was spawned out of the Hudson River. A total of fifty-eight fingerlings were made avail- able to the Hofstra Aquaculture Laboratory. The first batch of 35 was transported on 14 September 1995 to the campus facility where 30 were placed in a 100 gallon fiberglass tank with air and a powerhead water changer with a submersible pump hooked up to a trickle filter. Another five were placed in an aquarium with an underground filter. The ambient water temperature was 17°C and the pH was 7.2. while concentrations of ammonia and nitrites were negligible. Food was supplied by the hatchery. Biokiowa CIOO and C750 were fed at \9c body weight per day by hand. The light cycle was 12-hours light and 12-hours dark. Twenty-three finger- lings were brought to Aqualong Co. in Riverhead. New York on 7 December 1995 where they were maintained in a 6,000 gallon tankat47±°F. Fish in the 100-gallon fiberglass tank did not survive more than 2 days. Overfeeding may have been the cause. Fish held in the aquarium were cold-shocked on day 14 and only 2 survived. The 2 that survived grew from 55-mm to 1 lO-mm. similar to those at the hatchery which were kept in a raceway. The fish at Aqualong fared better; eleven of 23 survived the first snow storm and am- bient water conditions. Water continues to be changed weekly with no other care. Fish that survived seem to fare better on warm days after well water is exchanged and the temperature heats up to 54°F. They then began to feed on trout chow and became more active. Further growth stiidies will be conducted with those being raised in parallel at the New York Aquarium by Dr. Dennis Tho- ney. Journal of Shellfhh Research. Vol. 15, No. 2,465-532, 1996. ABSTRACTS OF TECHNICAL PAPERS Presented at the 88th Annual Meeting NATIONAL SHELLFISHERIES ASSOCIATION Baltimore, Maryland April 14-18. 1996 465 National Shellfisheries Association. Baltimore, Maryland Abstracts. 1996 Annual Meeting. April 14—18, 1996 467 CONTENTS APPLICATIONS OF BIOTECHNOLOGY TO SHELLFISH RESEARCH Eugene M. Burreson 101 Uses for the small subunit rihosomal RNA gene: Applications to Haplosporidium nelsoni 475 5. Craig Gary Waiter there is a bug in my clam; A molecular analysis of symbiont transmission in several marine bivalves 475 Thomas T. Chen, Jenn-Kan Lu, Standish K. Allen. Tomoyo Matsubara and Jane C. Burns Production of transgenic dwarf surfclams, Muliiua Uuercdis. with pantropic retroviral vectors 475 Rosemary Jagus Development of continuous marine invertebrate cell lines 476 Richard K. Koehn Biotechnology is a business, not a science 476 Kennedy T. Paynter Biotechnology and shellfish: Applying new scientific techniques to an old environment 476 James C. Pierce A bacteriophage P I high molecular weight genomic library from the oyster Crassostrea virginica 477 Dennis A. Powers. Vicky Kirby and Marta Gomez-Ghiarri Genetic engineering abalone: Gene transfer and ploidy manipulation 477 Kimberly S. Reece, John E. Graves and David Bushek Development of molecular markers for population genetic analysis of Perkinsus inannus 477 Gerardo R. Vasta Protein-carbohydrate interactions for self/non-self recognition in shellfish 478 BIVALVE FISHERIES John Aldred A profile of the East Hampton Town Shellfish Hatchery and reseeding program 478 W. S. Arnold, H. A. Norris and M. E. Berrigan Lease site considerations for hard clam aquaculturc in Florida 478 Jeffrey C. Brust, William D. DuPaul and James E. Kirkley Relative efficiency of 3.5" dredge rings in the offshore sea scallop fishery 479 T. Jeffrey Davidson and Gef Flimlin Shellfish management program — clam production 479 Christopher V. Davis and Sandra E. Shumway Larval and juvenile growth of Stimpson's surfclam — a new candidate species for aquaculturc development? 479 Bretl R. Dumbauld, Martin Peoples, David A. Armstrong and Stephen G. Ratchford A comparison of the effects of three different habitat modifications on intertidal clam populations in Pacific northwest coastal estuaries 480 William D. DuPaul, Robert A. Fisher and James E. Kirkley Natural and ex-vessel moisture content of sea scallops (Placopecten magellaiticiis) 480 Gef Flimlin Changes in hard clam, Menemuia mercenaria. fisheries of the mid-atlantic region in response to stock fluctuations. . 480 Alexander Gryska, G. Jay Parsons, Sandra E. Shumway, Kristin Geib, Ian Emery and Sue Kuenstner Polyculture of sea scallops suspended from salmon cages 481 Michael A . Rice The 1995 status of the shellfisheries for the northern quahog, Mercenaria mercenaria (L.) in New England 481 Jack M. Whetstone, William D. Anderson, Philip S. Kemp, Jr. and Randal L. Walker Development of the hard clam, Mercenaria mercenaria. fishery in the south atlantic region 481 BIVALVE SEED PRODUCTION Mark L. Homer. Robert Bussell and Chris Judy A cost-benefit evaluation of hatchery-produced oysters in Maryland 482 Richard C. Karney, Frank A. Dutra. David Dutra and Judy Dutra Hatchery and field culture techniques for the giant sea scallop Placopecten magellaniciis 482 Eric Powell, Susan Ford. John Klinck and Eileen Hofmann How important is the time of transplant in the success of oyster (relay) farming? 482 468 Abstracts, 1996 Annual Meeting, April 14-18, 1996 National Shellfisheries Association. Baltimore. Maryland Shawn M. C. Robinson, Jim D. Martin, Ross A. Chandler and Don Bishop A possible option for enhancement of the w ild fishery for the sea scallop. Placvpecten magellanicus 483 Mark Schexnayder. Randall Pausina. Ron Diigas and David Lavergne Selectivity and economic analysis of different cultch materials for oyster setting in Hackberry Bay. LA 483 CONSERVATION OF FRESHWATER MUSSELS Braven B. Beaty and Richard J. Neves Factors influencing the growth and survival of juvenile Villosa iris (Bivalvia; Unionidae) in an artificial stream system 483 Arthur E. Bogan Decline and decimation: The extirpation of the unionid freshwater bivalves of North America 484 David J. Berg, Sheldon I. Giittman and Emily G. Cantonwine Geographic variation in unionid genetic structure; Do management units exit? 484 Anne E. Keller Contaminant impacts on native freshwater mussels — lethal and sublethal responses relative to water quality criteria — 484 Anne E. Keller and Shane Ruessler Malathion toxicity to three life stages of unionid mussels 485 James B. Layzer The importance of habitat hydraulics m the restoration of native freshwater mussels 485 William A. Lellis and Connie S. Johnson Delayed reproduction of the freshwater mussel Elliptio complanuia through temperature and photoperiod control 485 Debbie C. Mignogno Freshwater mussel conservation and the Endangered Species Act 486 Richard J. Neves, Catherine Gatenby and Bruce Parker The exotic zebra mussel in North America: A dire prognosis for native freshwater mussels (Unionidae) 486 CRUSTACEAN BIOLOGY AND FISHERIES George R. Abbe and Cluney Stagg Some recent trends in Maryland blue crab populations 486 Peter G. Beninger, Annie Ferguson and Carole Lanteigne The gonopod tegumental glands of snow crab. Chionoecetes opilio: A closer look yields evidence for sexual function 487 Louis R. D'Abramo, Curtis G. Summerlin, William H. Daniels and H. J. Wan The effect of water volume and surface area of a culture containers on weight gain of juvenile freshwater prawns. Macrohrcuiuiiiu roseiihergi 487 Brett R. Dumbauld, David A. Armstrong, Kristine L. Feldman and John R. Skalski Field experiments on thalassinid shrimp control for oyster culture in Washington State 487 Cluney Stagg and George R. Abbe Relations among fixed station blue crab pot sampling results, reported Chesapeake Bay landings and winter dredge survey results 487 CRUSTACEAN HEALTH PROBLEMS Richard J. Cawthorn Impact of 'bumper car' disease on the North American lobster fishery 488 John A. Couch An overview of penaeid shrimp pathogens in U.S. waters 488 William S. Fisher, Patricia S. Glas, James R. Ray burn and Lee A. Courtney Toxicant effects on grass shrimp embryos 489 J. Frank Morado Histophagous ciliate diseases of Crustacea 489 Steve Re bach Effects of dimilin on the blue crab, Calliiiectes sapidus in shallow water habitats 489 Jeffrey D. Shields The parasitic dinoflagellates of marine crustaceans 489 National Shellfisheries Association. Baltimore, Maryland Abstracts. 1996 Annual Meeting. April 14^18, 1996 469 ECOLOGICAL FUNCTION OF BIVALVES Loren D. Coeit, Elizabeth L. Weniier, David M. Knott, Bruce W. Stender, Nancy H. Hadley and M. Yvonne Bobo Intertidal oyster reefs as critical estuarine environments: Evaluating habitat use, development and function 490 Jerry McConnick-Ray Oyster production — a large scale perspective 490 Elka T. Porter, Roger I. E. Newell and Lawrence P. Sanford Physical and biological scaling of benthic-pelagic coupling in coastal ecosystems; The role of bivalve suspension feeders 490 Michel Ropert, P. T. Goidletquer and J . P. Joly Trophic competition between the Pacific oyster Crassostrea gigas and the polychaete Lanice conchilega in the Bay of Veys ( France ) 49 1 ECONOMICS OF THE AQUACULTURE INDUSTRY Eric J. Powell Characteristics of on-bottom oyster rack structures in the Chesapeake Bay 491 LARVAL PROCESSES Sandra Brooke and Roger Mann Use of mesocosms for 'in situ' culture of marine invertebrate larvae 491 Margaret M. Dekshenieks, Eileen E. Hofman, John M. Klinck and Eric N . Powell Size and depth dependent larval mortality: A modeling study 492 Roger Mann Metapopulation dynamics of oysters in a subcstuary of the Chesapeake Bay: Estimating early life history mortality rates 492 Roger Mann and John Hainrick Metapopulation dynamics of oysters in a subestuary of the Chesapeake Bay: The role of physical transport 492 Marguerite C. Pelletier Genetic variation in time and size of metamorphosis in the bivalve, Muliiiiu lateralis 492 LOBSTER FISHERIES I and II Robert A. Bullis and Michael J. Syslo A rapid field-test for the detection of chemically stripped egg-bearing lobsters 493 Michael Clancy and J. Stanley Cobb Recruitment strategies in marine decapods: A comparative approach 493 Darrell Donahue, Robert Bayer, Terry Work and John Riley The effect of diet on weight gain, shell hardness, and flavor of new shell lobsters 493 Mary-Jane James-Pirri, J. Stanley Cobb and Richard A. Wahle Size and timing of settlement in postlarval lobsters: Is there a growth advantage? 494 Douglas S. Pezzack Overview of the Canadian lobster (Homanis aiucncainis) fishery: Recent trends m landings and management and the outlook for the future 494 John Riley and Gulni Ozbay Experiments to extend the survival of lobsters air shipped to distant markets 494 C. M. Rockel and W. H. Watson III A comparison of the osmoregulatory capabilities of coastal and estuarine lobsters 495 S. L. Waddy and D. E. Aiken Temperature control of recruitment in the American lobster 495 MOLLUSCAN DISEASE I Gustavo W. Calvo and Stephen J. Jordan Status of oyster diseases in Maryland's oyster recovery areas 495 Lisa M. Calvo, Eugene M. Burreson. Christopher F. Dungan and Bob S. Roberson Perkinsus inarinus transmission dynamics in Chesapeake Bay 496 470 Absiracis. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore. Maryland Nancy H. Hadley, M. Yvonne Bobo, Donnia Richardson, Loren D. Coen and David Bushek Use of specific-pathogen-tree (SPF) oysters to measure growth, mortality, and onset of MSX and dermo disease in South Carolina 496 Dale S. Mulholland and Frank E. Friedl Distribution and population dynamics of a hydrozoan inquiline symbiont of the eastern oyster 496 Leah M. Oliver, W. S. Fisher, E. M. Burreson, L. M. Ragone-Calvo, S. E. Ford and J. Gandy Perkinsus mariiuis tissue distribution and seasonal variation in oysters {Cnissoslrea virginica) from Florida. Virginia and New York 497 Kimberly S. Reece, Mark E. Siddall and John E. Graves Phylogenetic analysis of the genus Perkinsus based upon actin gene sequences 497 Bob S. Roberson. Tong Li, Christopher F. Dungan and Eugene M. Burreson Perkinsus marinus. flow cytometric immunoassay and intcrannual abundance in Chesapeake Bay estuaries 497 Jeffrey D. Shields, Frank O. Perkins and Carolyn S. Friedman Hematological pathology of wasting syndrome in black abalone 498 Nancy A. Stokes, Juanita G. Walker and Eugene M. Burreson Comparison of HaplosporiJium nelsoni diagnostic techniques: Polymerase chain reaction outperforms histology 498 MOLLUSCAN DISEASE II Cal Bair-Anderson and Robert S. Anderson The effect of pentachlorophcnol on NADPH production in oyster hemocytes: Immunomodulatory consequences 498 David Brown, George Clark and Rebecca \'an Beneden Isolation of a cDNA clone from Menenarui mercenarui that codes for a protein related to the cytochrome P450 111 subfamily of enzymes 499 David Bushek, Susan E. Ford and Marnita M. Chinlala Infectivity and pathogenicity of Perkinsus marinus. 3. Fecal elimination 499 Marnita M. Chintala, David Bushek and Susan E. Ford Infectivity and pathogenicity of Pcrkuisus mannus. 1 . Parasite characteristics 499 Christopher F. Dungan, Rosalee M. Hamilton, Eugene M. Burreson and Lisa M. Ragone-Calvo Identification of Perkinsus marinus portals of entrv' of histochemical immunoassays of challenged oysters 500 Mohamed Faisal, Jerome F. LaPeyre and Craig D. Wright Protease blockers inhibit Perkinsus marinus in vitro and in vivo 500 Susan E. Ford, Marnita M. Chintala and David Bushek Infectivity and pathogenicity of Perkinsus marinus. 1. Dosing methods and host reponse 500 Frank E. Friedl Oysters, oxygen metabolism, and hemocytes 501 Jerome F. LaPeyre, Kathleen A. Garreis, Heather A. Yarnall and Mohamed Faisal Emerging evidence of extracellular proteases as important virulence factors of Perkinsus marinus 501 MOLLUSCAN FEEDING STUDIES Andrew G. Bauder, Allan D. Cembella, Michael A. Quilliam and Jon Grant Kinetics of diarrhetic shellfish toxins in the bay scallop. Argopecten irraclians 501 Peter G. Beninger and Suzanne C. Dufour Mucocyte distribution and relationship to particle transport on the pseudolamellibranch gill of Crassostrea virginica ( Bivalvia; Ostreidae) 502 V. Monica Bricelj, David Laby and Allan D. Cembella Differential sensitivity and PSP toxin accumulation in two clam species. Spisula soiidissima and Mya arenaria 502 Peter J. Cranford, Barry T. Hargrave and Conrad A. Pilditch Temporal perspectives on feeding and digestion by suspension-feeding bivalves 503 Jon Grant and Conrad A . Pilditch Field and modelling studies of bivalve culture in a boreal environment 503 i hris J. Langdon and Mike A. Buchal Delivery of low molecular weight, water-soluble nutrients to marine suspension feeders 503 National Shellfisheries Association. Baltimore. Maryland Abstracts. IW6 Annual Meeting. April 14-18. 1996 471 Bruce A. MacDonald, J. Evan Ward and Gregory S. Bacon Feeding activity in the sea scallop Placopecten magelUmuus: Comparison of field and laboratory data 503 Carter R. Newell In situ measurements of mussel {Mxtihis etiulis) energy acquisition in relation to seston concentration in a subtidal Maine estuary: How important is the shell gape response? 504 C. A. Pilditch, J. Grant. A. L. Mallet, C. E. A. Carver and P. J. Cranford Seston supply to scallops in suspended culture 504 Gerard Thouzeaii, Erederic Jean and Yolanda Del Amo Sedimenting phytoplankton as a major food source for suspension-feeding queen scallops (Aequipecten opercularis L. ) off Roscoff ( western English Channel I'!" 504 G. H. Wikfors. B. C. Smith, J. H. Alix and M. S. Dixon When is it time to feed the scallops? 505 MOLLUSCAN NUTRITION Brian Bayne and Tony Hawkins Feeding behaviour at high and variable seston loads 505 S. Craig Cary Obligate endosymbiotic associations between chemoautotrophic bacteria and marine bivalves; Nutritional implications to the early life stages 505 Fu-Lin E. Chu Lipid nutrition and fatty acid synthesis in oysters 506 S. M. Gallager Ciliary suspension-feeding and particle selection in mollusc larvae 506 Daniel A . Kreeger and Roger I. E. Newell Omnivory by the mussell. Ceukensia demissa 506 Philippe Soudant, Yanic Marty, Jean Rene LeCoz, Jeanne Moal and Jean Francois Samain How are bivalve broodstock and larvae adapted to meet their nutritional requirements for lipids 507 J. Evan Ward, Jeffrey Levinton, Sandra Shumway and Terri Cucci Looking into the "black box": Feeding strategies and limitations of suspension-feeding bivalves 507 MOLLUSCAN REPRODUCTION James W. Anderson and Richard K. Wallace Induction of triploidy in unconditioned eastern oysters. Crassoslrea virgintca. using nitrous oxide under increased pressure 507 Brian F. Beal, Stephen R. Fegley and K. W. Vencile Recruitment patterns of Myci arenaria L. from eastern and southwestern Maine: 11. Effects of site, tidal height, and predator exclusion 508 Stephen R. Fegley, Brian F. Beal and K. W. Vencile Recruitment patterns oi Mya arenaria L. from eastern and southwestern Maine: I. Short-term effects of site, tidal height, and predator exclusion 508 Dan C. Marelli, William S. Arnold, Catherine Bray and Melissa Harrison Estimates of recruitment and adult abundance in three Florida populations of bay scallops (Argopecten irradians) 509 Francis X. O'Beirn and Randal L. Walker Variations in gametogenesis and sex ratios in oysters along an intertidal gradient 509 John E. Supan and Charles A. Wilson Analyses of gonadal cycling by oyster broodstock, Crassostrea virginica (Gmelin), in Louisiana 509 John E, Supan, Charles A. Wilson and Standish K. Allen, Jr. The effect of salinity change on the synchrony of polar body development in fertilized oyster eggs (Crassostrea virginica [Gmelin]) 509 John E. Supan, Charles A. Wilson and Standish K. Allen, Jr. The effect of Cytochalasin B (CB) dosage on the survival and ploidy of Crassoslrea virginica (Gmelin) larvae in Louisiana 510 Karin A. Tammi, Michael A. Rice, Wayne H. Turner and Bethany A. Starr A comparison of artificial spat collectors in the Westport River. MA 510 472 Absiracts. 1996 Annual Meeting, April 14-18. 1996 National Shellfisheries Association. Baltimore, Maryland Janzel R. Villalaz Histological study of reproduction in Argopeclen ventricosus 510 MARINE GENETICS Patrick M. Gaffney and Francis X. O'Beirn Nuclear DNA markers for Cnissosireu species identification 510 Ximing Guo, Dennis Hedgecock, William K. Hershberger, Kenneth Cooper and Standish K. Allen, Jr. Genetics of sex determination in Crossoslrea oysters: A single locus model 511 Dennis Hedgecock Hybrid vigor is pervasive in crosses among inbred lines of Pacific oysters 511 Gang Li and Dennis Hedgecock Mitochondrial DNA variation within and among larval cohorts of Pacific oyster, Crassostrea gigas. detected by PCR-SSCP analysis 511 Suifen Lyu, Standish K. Allen, Jr., Gregory A. Debrosse and Patrick M. Gaffney Attempted hybndization of eastern and Pacific oysters using bridging crosses 512 Daniel J. McGoldrick and Dennis Hedgecock Microsatellite marker development in the Pacific oyster {Crassostrea gigas): Variability, transmission, linkage and QTL mapping 512 P. D. Ramon and T. J. Hilbish The role of phylogenetic distance on the disruption of doubly uniparental mtDNA inheritance in h\ bnd mussel (Mxtilus) populations 512 John Scarpa, Leslie Stunner, Everette Quesenberry, Ross Longley and David Vaughan Performance of triploid oysters, Crassostrea virginica. grown by project O.C. E.A.N, participants 512 Jeffrey R. Wakefield and Patrick M. Gaffney DGGE reveals additional population structure in American oyster {Crassostrea virgiiuca) populations 513 Ami E. Wilbur and Patrick M. Gaffney Geographic variation in morphology of the bay scallop, Argopeclen irradians (Lamarck) 513 OYSTER DISEASE RESEARCH PROGRAM Hafiz Ahmed and Gerardo R. Vasta Glycosidases in Perkinsus mannus: Purification and characterization of P-D-glucosidase 513 Standish K. Allen, Jr., Ximing Guo, Gene Burreson and Roger Mann Heteroploid mosaics and reversion among triploid oysters. Crassostrea gigas: Fact or artifact 514 Ryan B. Carnegie, Bruce J. Barber and Christopher V. Davis Growth and timing of juvenile oyster disease (JOD)-induced mortality of Crassostrea virginica in the Damariscotta River. ME, USA 514 Fu-Lin E. Chu, Aswani K. Volety and Georgetta Constantin Intracellular and extracellular lysosomal enzyme activities in eastern oysters (Crassostrea virginica) 514 Gregory A. Debrosse and Standish K. Allen, Jr. Cooperative regional oyster selective breeding (CROSbreed) project 514 C, Austin Farley, Earl J. Lewis, David Relyea, Joseph Zahtila and Gregg Rivara Resistance studies for juvenile oyster disease (JOD) 515 Julie D. Gauthier and Gerardo R. Vasta Inhibition of Perkinsus marinus in vitro proliferation by heterologous plasma 515 Ximing Guo, Standish A. Allen, Jr. and Patrick M. Gaffney Gene transfer through hybrid partial gynogenesis between the Pacific and American oysters 515 Earl J. Lewis, C. Austin Farley, Ana Baya and Eugene B. Small Juvenile oyster disease — transmission and bacteriological studies 516 Adam Marsh, Anita C. Wright and Gerardo R. Vasta Isolation and characterization of marker genes for Perkinsus mannus 516 Donald Meritt, Kennedy T. Paynter and Robert Pfeiffer Reconstruction of a natural oyster bar in the Choptank River using hatchery produced oyster seed 516 Kennedy T. Paynter, Patrick M. Gaffney and Donald Meritt Evaluating eastern oyster stocks for resource rehabilitation 517 National Shellfisheries Association. Baltimore. Maryland Ahsiracts. 1996 Annual Meeting. April 14—18. 1996 473 Jose Antonio F. Rohledo, Adam G. Marsh, Anita C. Wright and Gerardo R. Vasta Assessment of geographic variability in Perkinsiis miinmis 517 Aswani K. Volety and Fu-Lin E. Chit Acid phosphatase: A virulence factor of the protistan parasite. Perkinsiis mariiius against host oyster's defense? 517 SHELLFISH NEOPLASIA y\f. S. Arnold. T. M. Bert, D. M. Hesselman and N. J. Blake Gonadal neoplasia in hard clams (Meixciuiria spp, ) from the Indian River lagoon. Florida 518 Bruce J. Barber Gonadal Neoplasms in Mya arenana: What do we know? 518 Arnold G. Eversole and Peter B. Heffernan Gonadal neoplasia in northern and southern quahogs and their hybrids in South Carolina 518 Shawn M. McLaughlin, C. Austin Farley and Christopher C. Judy An update on softshell clam {Mxa tireiuirici) sarcoma in the Chesapeake Bay and the declining fishery 519 J. Frank Morado and Donald V. Lightner The rare occurrence of neoplasia in Crustacea; Myth or sampling artifact 519 Mary-Susan Potts Effects of hematopoietic neoplasia on reproduction and population size distribution in the soft-shell clam 519 Roxanna Smolowitz Neoplasia and other pollution associated lesions in Mya aremina from Boston Harbor 520 R. J. Van Beneden, L. R. Rhodes, D. J. Brown and G. R. Gardner Investigation of molecular mechanisms of tumorigenesis in bivalve gonadal tumors 520 Charles W. Walker, Sharon A. Key, Joseph E. Mulkern, Shalini Verma and Jocelin A. Jacobs Expression of the tumor suppressor gene. p53 in normal and leukemic clam blood cells in vivo and in vitro 520 STOCK ASSESSMENT James G. Boyd, William D. Anderson and Guy M. Yianopoulos Using real time data with a PC-based GIS for shellfish management 521 Stephen R. Fegley, Susan E. Ford, John N. Kraeuter, David R. Jones and Harold H. Haskins Relative effects of harvest and disease mortality on eastern oyster populations in Delaware Bay 52 1 Mark L. Homer, Mitchell Tarnowski and Lisa Baylis Eastern oyster stock assessment in Maryland 52 1 G. F. Smith and K. N. Greenhawk Morphological differentiation of the fringing and patch oyster reef types in Chesapeake Bay: A comparative evaluation 522 Mitchell L. Tarnowski, Mark L. Homer, Lisa Baylis and Robert Bussell Molluscan inventory of Maryland's coastal bays 522 WATER QUALITY AND GOVERNMENT REGULATION Paul G. Comar Control of Vibrio vulnificus growth to reduce risk in shellfish consumption 522 Elizabeth Fellows Assessing water quality: New directions 523 Paul Orlando, John Klein, Daniel Farrow, Anthony Pait, Dorothy Leonard and Jamison Higgins Targeting strategies for shellfish restoration in the Gulf of Mexico: Results of a regional strategic assessment process 523 G. /. Scott, M. H. Fulton, D. Porter, S. Strozier, E. D. Strozier, P. B. Key and J. W. Daugomah The effects of urbanization on the American oyster, Crassosrrea virgiiiiin (Gmelin) 523 William D. Watkins Synopsis of FDA research related to water quality 524 POSTER SESSION Susan B. Athanas and David B. Rouse Incidence of fouling at two mariculture sites in Bon Secour Bav. Alabama 524 474 Absinuts. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore, Maryland P. Baker and D. J. Hornbach Can a species be "fouled"' into extinction'^ Zebra mussels vs native bivalves in the upper Mississippi River 524 Kerri M. Benlkowski, J. Evan Ward and Roger I. E. Newell Paddles or sieves: Testing the mechanisms of particle retention in bivalves 524 M. Yvonne Bobo, Donnia Richardson, Thomas C. Cheng, Elizabeth McGovern and Lore n Caen Seasonal cycle of Haplo.spondiitm nelsoiii (MSX) in intertidal oysters. Crassostrea virginka. in South Carolina 525 Guy W. Bruni and Yolanda J. Brady Effects of PeikinsKs mariiui.s on cultured Mobile Bay oysters 525 Gregory P. Died Predator- and prey-differentiated repair frequencies in the moon snails Euspira heros and Neveriia ditpUcaui versus the whelks Biisxcon caiica and Biisxcon canalicidalum from Cape May County. New Jersey 525 Ray Grizzle and Mike Castagna Spatial patterns of intertidal oyster reefs in the Canaveral National Seashore. Florida 526 Peter L. Haaker, Gary E. Davis and Ian K. Taniguchi Serial depletion in marine mvcrtcbratc diving fisheries 526 D. J. Hornbach. T. Deneka and P. Baker Federally endangered freshwater mussels in the St. Croix River: Microhabitat and mussel community associations 526 Dorset H. Hurley and Randal L. Walker The effects of larval stocking density on growth and survival of laboratory reared SpisuUi solidissinui similis 527 V. S. Kennedy, M. Asplen and T. Hall Salinity effects on mtroduced drcisscnid mussels 527 Tim L. King. Mary E. Smith, Rita F. Villella, Priscilla I. Washington and David A. Weller Genetics and systematics of freshwater mussel species: a tissue repository 527 Tim L. King. Rita F. Villella, Mary E. Smith and Michael S. Eackles Discontinuity in the genetic population structure of the green floater Lasmigona suhviridis 527 David P. Lemarie. David R. Smith, Rita F. Villella and David A. Weller Evaluation of tag types and adhesives for marking freshwater mussels 528 A. D. McKinney, R. G. Hudson and Margaret L. Barfield Species specificity and effect of pH on the response of freshwater mussel juveniles to acute copper toxicity 528 Carter R. Newell and Bohdan M. Slabyj Fate of potential bacterial contaminants as a function of contact surface in shellfish wet holding tanks 528 Elizabeth A. Orbacz. Ami E. Wilbur, Jeffrey R. Wakefield and Patrick M. Gaffney RFLP analysis of genetic diversity in a Siberian population of the Japanese scallop (Patinopecten yessoensis) 529 F. Scott Rikard, Richard K. Wallace and Christopher L. Nelson Management strategies for fouling control in Alabama oyster culture 529 Melanie K. Shadoan and Ronald V. Dimock, Jr. Sensory physiology of Glochidia larvae of the freshwater mussels Utterbackia imbecilUs and Megulonais nervosa 529 Lioudmila V. Spektorova The culture variability of Monochiysis liilheri as an advantage for shellfish culture 530 Jeffrey J. Springer, JoAnn Burkholder and Sandra E. Shumway Effects of the toxic dinotlagellate. Pfieslehci piscicuhi. on juvenile bay scallops (Argopecten irmdians. Lamarck) 530 R. A. Tankersley, J. J. Hart and M. G. Wieber Developmental shifts in the feeding biodynamics of juvenile Utterbackia imbecilis (Mollusca: Bivalvia) 530 Wayne H. Turner, Karin A. Tammi and Bethany A. Starr What a year to be a mud crab! Three years on the bay scallop restoration project. Westport River estuary. Massachusetts 531 Randal L. Walker, Dorset H. Hurley and Michelle L. Jansen Fecundity estimates of the southern surfclam. Spisula solidissinui similis 531 Charles W. Walker and Michael P. Lesser Prepared food coupled with manipulation of photoperiod yield an out-of-season crop for the northeastern sea urchin.. . 531 Ji'nnifer Wojcik and Kennedy T. Paynter Myeloperoxidase activity from blood cells of the eastern oyster. Crassostrea virginica 532 National Shellfisheries Association, Baltimore. Maryland Abslnicl. 1996 Annual Meeting. April 14-18, 1996 475 APPLICATIONS OF BIOTECHNOLOGY TO SHELLFISH RESEARCH 101 USES FOR THE SMALL SUBUNIT RIBOSOMAL RNA GENE: APPLICATIONS TO HAPLOSPORIDWM NEL- SONI. Eugene M. Burreson,* Virginia Institute of Marine Sci- ence, College ot William and Mary. Gloucester Point. VA 23062. The small subunit ribosomal RNA (SSU rRNA) gene is a part of the rRNA transcription unit, which is present m 100s to 1000s of copies within the genome. A large part of the SSU rRNA gene sequence is well conserved across all eukaryotes, however there are also hypervariable regions that are species-specific and these areas can serve as targets for molecular probes and primers. We isolated genomic DNA from HaplospDiulium nelsoni and ampli- fied the SSU rDNA via the polymerase chain reaction (PCR) with eukaryotic universal primers. The gene was sequenced and two species-specific regions were identified to use as a DNA probe and PCR primers for sensitive and specific detection of H . nelsoni. The DNA probe has been used in in situ hybridizations for H. nelsoni diagnosis in histological samples and the PCR primers have been used for detection of H. nelsoni from infected oyster tissue or hemolymph. We plan to use the probe and primers for elucidation of the H. nelsoni life cycle. With the use of both the probe and primers, we have determined that H nelsoni was intro- duced to the cast coast of the U.S. by importations of infected C. gigas. The amplification products from haplosporidian- infected C. gigas were sequenced using the PCR primers and. except for one transition, were identical to the H. nelsoni SSU rDNA sequence. Comparison of the sequence of conserved regions of the SSU rRNA gene across taxa allows inference of phylogenetic relation- ships. We have sequenced this gene in several other haplosporid- ians. Multiple alignment and phylogenetic analysis with ciliates. dinoflagellates, and apicomplexans showed that the phvlum Hap- lospondia has an alveolate ancestry. Within the Haplosporidia, these analyses showed that the genus Haplosporidium. as pres- ently defined, is not monophyletic. WAITER THERE IS A BUG IN MY CLAM: A MOLECU- LAR ANALYSIS OF SYMBIONT TRANSMISSION IN SEV- ERAL MARINE BIVALVES. S. Craig Cary, Graduate College of Marine Studies. University of Delaware, Lewes. DE 19958. The power of molecular diagnostic tools has recently been applied to certain areas of shellfish research. Nucleio acid probe technology based on I6S rRNA sequences provided the high res- olution necessary to identify the presence and location of ex- tremely low numbers of bacteria in eukaryotic cells. Ribosomal genes offer ideal targets for hybridization probes because, a) they are present in multiple copies, b) they offer a range of variable regions, including some regions that are invariant among all living organisms, and others that are unique to particular organisms or related groups of organisms. In addition, actively growing bacte- rial cells may contain as many as 104 ribosomes each a potential hybridization target for a complementary oligodeoxynucleotide probe. Oligodeoxynucleotide probes directed against rRNA targets are rapidly becoming one of the most powerful techniques in mi- crobial ecology enabling the detection of ribosomal genes at very low copy number from a mixed natural population at defined lev- els of phylogenetic specificity. These probes can be utilized in both standard detection analysis using the Polymerase Chain Re- action or for the actual localization of the bacteria within the host organism using highly sensitive in situ hybridization protocols. Collectively these methods provide highly sensitive detection di- agnostic capabilities. Recently these methods have been applied to determine the symbiont transmission mechanism in several bi- valves inhabiting deep sea hydrothermal vents and cold seep hab- itats. The resolution has provided insight into the mechanism of symbiont transmission and constraints on larval dispersal and set- tlement. PRODUCTION OF TRANSGENIC DWARF SURFCLAMS, MULINIA LATERALIS, WITH PANTROPIC RETROVIRAL VECTORS. Thomas T. Chen,* Biotechnology Center. Depart- ment of Molecular and Cell Biology. University of Connecticut. Stoors. CT; Jenn-Kan Lu, Department of Biological Sciences. University of Maryland at Baltimore County. Baltimore. MD; Standish K. Allen, Haskin Shellfish Research Laboratory, Institute of Marine and Coastal Sciences. Rutgers University. Port Norris. NJ; Tomoyo Matsubara, Department of Pedia- trics, UCSD School of Medicine, La Jolla. CA; Jane C. Burns, Department of Pediatrics. UCSD School of Medicine. La Jolla. CA. A pantropic pseudotyped retroviral vector containing the enve- lope protein of vesicular stomatitis virus was used as a gene trans- fer vector in the dwart^ surfclam. Mulinia lateralis. These pan- tropic retroviral vectors have an extremely broad host cell range and can infect many non-mammalian species. Newly fertilized dwarf surfclam eggs were electroporated at 700 volts in the pres- ence of I X 10"* cfu of pantropic pseudotyped retroviral particles. Infection was well tolerated and did not affect the survival rate of the embryos. Gametes collected from P, presumptive transgenic animals were analyzed for the presence of provirus by PCR, and in different experiments 13-337^ of the gamete pools were positive for the transgene. Dot blot hybridization of DNA samples from the F| offspring of two different crosses between infected P, and wild type individuals revealed that 28% and 31% of F, offspring were transgenic, respectively. Southern blot analysis of DNA isolated from PCR-positive F, animals confirmed integration of a single copy of the provirus into the host genome. Thus, the germlines of these two P, transgenic animals were mosaic for the transgene. Expression of b-galactosidase encoded by the provirus was de- tected in transgenic but not control surfclam embryos. Pantropic 476 Abstract. 1996 Annual Meeting, April 14-18. 1996 National Shellfisheries Association, Baltimore, Maryland pseudotyped retroviral vectors provide a useful method for the stable introduction of foreign genetic information into surfclams and may facilitate the introduction of desirable genetic traits into commercially important shellfish and crustaceans. DEVELOPMENT OF CONTINUOUS MARINE INVERTE- BRATE CELL LINES. Rosemary Jagus,* Center of Marine Biotechnology, Suite 236 Columbus Center, 701 E. Pratt Street, Baltimore, MD 21202. We are attempting to develop continuous marine invertebrate cell lines by transfecting sea urchin disrupted gastrula cells with oncogenes of various types. We propose to use the broad host range pseudotyped retroviral vector that has been successfully used in finfish and the dwarf surf clam. Our overall strategy is to transform disrupted gastrula cells with a range of oncogenes such as those for myc. ras. SV40 large T-antigen. nonfunctional PKR, and elF4E, using the pseudotype retroviral vector. pLGRNL. In this Moloney murine leukemia virus-based vector, the virus coat protein is replaced by the VSV-G protein giving rise to the vector's characteristic wide host range. The vector also contains the neo- mycin phosphotransferase gene to allow selection of stable trans- formants. Our immediate goals are: a) to construct selectable pseudotype vectors expressing growth promoting genes under the control of a suitable sea urchin promoter; b) demonstrate expres- sion in short term disrupted gastrula cultures; c) select neomycin resistant colonies. Cells of the sea urchin, StrongyUxentnitiis purpwatus . will he used to demonstrate the approach since there is a relative abun- dance of information on sea urchin molecular genetics, including the characterization of strong promoter elements. The approach will provide a strategy for the development of immortalized cell lines from other economically and therapeutically important ma- rine invertebrates. We have worked initially on Phase a). Because genes in this vector will integrate into the host DNA, and because we will be transforming with oncogenes, there is a potential hu- man health risk. Consequently, we are first assessing the ability of several sea urchin promoters to support transcription in human cells (HeLa). We will begin our studies with promoters that func- tion well in sea urchin cells but poorly in human cells. We have engineered sea urchin promoters into the mammalian expression pCDNAl/neo containing the reporter gene chloramphenical ami- noacyl transferase (CAT). Of the sea urchin promoters available, we have chosen those that are expressed uniformly with respect to cell type and which exhibit activity well into gastrulation. These include SpHE, the hatching enzyme promoter, Spec-1, and sub- regions of the Cyllla (actin) promoter. Currently, the abilities of these promoters is being assessed for transient expression in HeLa cells. Our next stage will be to detemiine the stability of the pseudotype vector to a range of sea water concentrations. This .■•;;ige does not require sea urchin cells, since we can assess virus viability in HeLa cells. BIOTECHNOLOGY IS A BUSINESS, NOT A SCIENCE. Richard K. Koehn. University of Utah, Salt Lake City. UT 84112. While scientific research is the fuel of technological innovation and economic growth, the commercial exploitation of scientific discovery is neither straight forward nor simple. Biotechnology is an industry with high economic volatility reflecting the rapidly changing fortunes of a new industry in both rapid growth and consolidation. The climate for investment m biotechnology has been described as having moved from passion to panic. If agricultural biotechnology is a step-child of medical biotech- nology, then marme biotechnology is the orphan. Marine research has the potential to be of economic value, but to date marine biotechnology has been characterized more by poetic sweeps than by concrete economic growth. The reasons for this will be dis- cussed. Most marine research is performed in universities or other non- profit organizations. As such, the policies of these institutions on conflict-of-interest, equity position, royalty sharing, management of intellectual property, etc., are critical to the successful com- mercialization of marine science. How these relate to the financial, legal, managerial, and scientific aspects of marine biotechnology will be discussed. If marine research is to fuel significant growth of marine bio- technology, the research must become much more market driven. Biotechnology is not just molecular biology research, but science in service of technological innovation and commerce. BIOTECHNOLOGY AND SHELLFISH: APPLYING NEW SCIENTIFIC TECHNIQUES TO AN OLD ENVIRON- MENT. Kennedy T. Paynter.* Department of Zoology, Univer- sity of Maryland, College Park, MD 20742. The advent of molecular biology, the manipulation of RNA, DNA and associated molecules, has revolutionized the scientific world in the last decade, it is now possible to add specific genes to the genome of almost any species, to detect a single abnormal cell among tens of thousands of normal ones, and to determine genetic identity with great accuracy. Needless to say, the medical community wasted no time in embracing these advances and ap- plied them to detect and/or cure a variety of diseases. At present, these techniques are also being applied in the agriculture industry to improve production and the quality of farmed products. While these advances will likely result in direct improvement of indus- trial output, the application of molecular tools to conduct research on important ecological or environmental issues has been more slow. There are a few examples of molecular techniques being em- ployed in marine research 10 to 15 years ago but. for the most part, molecular techniques have been commonly employed by marine scientists only in the last few years. However, the promise of molecular techniques to marine research is great. For aquaculture, the construction of fast-growing transgenic fish and shellfish might National Shcllfisheries Association. Baltimore. Maryland Abstract. 1996 Annual Meeting. April 14-18. 1996 477 vastly improve production rates. Comparison of various segments of DNA among aquatic species could improve detection capabil- ities for important pathogens by several orders of magnitude. This might allow for the detection of unknown life cycle stages of certain parasites, clarify relationships between disparate groups of pathogens and establishing phylogenetic relationships between pathogens. Most importantly these tools will enable us to better understand the biology and ecology of marine and estuarine species. A BACTERIOPHAGE PI HIGH MOLECULAR WEIGHT GENOMIC LIBRARY FROM THE OYSTER, CRASSOS- TREA VIRGIMCA. James C. Pierce,* Department of Biologi- cal Sciences. Philadelphia College of Pharmacy and Science. Phil- adelphia, PA 19104-4495. The bacteriophage PI cloning system is able to generate bac- terial clones with insert sizes up to 100 kb. These clones have a number of unique features including, positive selection, single copy plasmid replication and a inducible high copy number repli- con. Rare cutting restriction sites and T7 and Sp6 promoters bor- der the cloning site and facilitate analysis and characterization of PI clones. My laboratory is currently constructing a PI genomic library from the eastern oyster, Crassostrea virginica. To mini- mize DNA contamination from algal and microbial cells we have obtained relatively pure sperm cells from male gonads by direct puncture with a capillary tube. High molecular weight (HMW) genomic DNA in the megabase (Mb) range was isolated by cell lysis and purified by sucrose gradient centrifugation. Genomic inserts were generated by Sau3AI partial digest followed by su- crose gradient fractionation. PI clones containing HMW oyster genomic inserts are being constructed using a two stage, in vitro PI packaging reaction. Our goal is to construct a genomic library that contains 20.000 individual PI clones with an average insert size of about 80 kb. Since the C. virginica haploid genome is estimated to be 3.8 x 10" bp. a genomic library of this size will give a library coverage of approximately 4 fold. Specific clones are isolated using PCR screening of clone pools and by a nonra- dioactive colony purification protocol. A PI library for C. virgin- ica will significantly improve our ability to perform genetic phys- ical mapping studies and in the isolation and study of specific genomic regions. PI clones have proven to be excellent substrates for genome targeting in mammalian systems and may prove useful in the genetic engineering of the oyster genome. Construction of transgenic oysters which have novel and commercially important traits will be facilitated by access to a PI oyster library. GENETIC ENGINEERING ABALONE: GENE TRANSFER AND PLOIDY MANIPULATION. Dennis A. Powers,* Vicky Kirby, and Marta Gomez-Ghiarri, Hopkins Marine Station. Stanford University. Pacific Grove. CA. We are genetically engineering abalone with enhanced growth by gene transfer and ploidy manipulation. We have: (i) developed efficient methods for simultaneously transferring genes into thou- sands of abalone eggs, (ii) cloned the first abalone promoter, (iii) coupled various promoter to reporter genes and the coho salmon growth hormone, (iv) transferred these constructs into abalone, (v) determined integration and expression, and (vi) developed triploid abalone with enhanced growth. The abalone P-actin promoter was cloned and sequenced. This promoter and others were coupled to luciferase, p-galactosidase and coho salmon growth hormone. These recombinant plasmids were linearized and introduced in fertilized eggs of the red abalone [Haliatis rufescens) by electro- poration. When the conditions were optimized, the majority of the embryos became transgenic and retained the constructs for more than a year. Southern hybridization analyses suggested head-to-tail concantermers integrated in the genome and these genes were expressed. Since it takes several years to reach sexual maturity, the transmission of these transgenes to the next generation is being evaluated. In addition to our transgenic work, we have also used pressure, temperature and chemical treatment to manipulate the ploidy of abalone. Although these treatments had different effi- ciencies and the results varied depending upon time after fertiliza- tion, we have successfully generated triploid abalone that grow significantly faster than their diploid counterparts. We are using triploid manipulation of transgenic abalone to create unique strains of abalone for aquaculture purposes. DEVELOPMENT OF MOLECULAR MARKERS FOR POP- ULATION GENETIC ANALYSIS OF PERKINSUS MARI- NVS. Kimberly S. Reece* and John E. Graves, Virginia Insti- tute of Marine Science, College of Willima & Mary. Gloucester Point, VA 23062; David Busliek, Barueh Marine Field Labora- tory, University of South Carolina. Georgetown. SC 29440. We are investigating the population structure of the oyster pathogen Perkinsus marinus to help identify mechanisms of dis- persal and migration. Infected oysters collected from Connecticut to Texas have been used to produce in vitro cultures of the parasite for genetic analysis. We have constructed a genomic library, iden- tified regions of the DNA which show intra-specific variability, and designed primers for these regions. Using universal primers for polymerase chain reactions (PCR). fragments of P. marinus actin, I8S rRNA, the internal transcribed spacer (ITS) region of rRNAs, ATPase 6 (mitochondrial) and serine protease genes were amplified and cloned. Amplified gene fragments were labeled with digoxigenin and used as probes to screen the genomic library for lambda phage clones containing the genes and flanking sequences (avg. insert = 18 kbp). Phage clones containing the genes were isolated and the inserts containing P. marinus DNA were mapped with restriction enzymes. Regions flanking the genes were sub- cloned into plasmid vectors, sequenced, and non-coding regions identified by computer searches of the DNA sequence for coding regions. Based upon these results. PCR primers were designed to amplify 1-3 kbp non-coding fragments of the flanking regions. DNA isolated from geographically distinct P. marinus cultures was amplified using the newly designed primers and surveyed with a suite of restriction endonucleases to assess fragment length poly- 478 Abstract. 1996 Annual Meeting. April 14-18, 1996 National Shellfisheries Association. Baltimore, Maryland morphisms between isolate cultures (RFLP analysis). Genetic variation has been observed at several loci. Many of the fragments are also being sequenced to assess variation at the nucleotide level. As more isolates become available, RFLP and sequence analysis will continue and both phenetic and cladistic analyses will be em- ployed to determine relationships among isolates from different areas. PROTEIN-CARBOHYDRATE INTERACTIONS FOR SELF/NON-SELF RECOGNITION IN SHELLFISH. Ger- ardo R. Vasta,* The Center of Marine Biotechnology, University of Maryland Biotechnology Institute, Columbus Center, 701 E. Pratt St.. Baltimore. MD 21202. Viral, bacterial, fungal and protozoan epizootic diseases are recognized as significant detrimental factors for the successful exploitation of natural and cultivated stocks of marine shellfish, such as crabs, shrimps, oysters and clams. Because shellfish do not synthesize immunoglobulin antibodies, those established methods for the control of disease in vertebrates, such as rational vaccination programs, can not be applied to moUuscan and crus- tacean species. Therefore, the elucidation of their internal defense mechanisms in shellfish is of utmost importance. In the long term, the identification and thorough characterization of structure and function of the recognition/effector gene products and the detailed understanding of how their expression is regulated, will enable us to enhance disease resistance through their adequate stimulation and to further apply transgenic approaches to the development of disease-resistant shellfish species. The participation of certain car- bohydrate-binding proteins (lectins) in non-self recognition/ defense mechanisms is supported by ample evidence obtained in a variety of animal models, including man. Within the invertebrates, at least three major lectin categories can be identified: one group includes lectins that show significant homology to membrane- integrated or soluble C-type vertebrate lectins, such as homo- logues of the mannose-binding receptor. The second includes P-galactosyl-specific lectins homologous to the S-type vertebrate lectins (Galectins). The third group is constituted by lectins that show homology to acute phase reactants from vertebrates, such as C-reactive protein and serum amyloid P. The multiplicity in lectin specificity and the nature and distribution of the carbohydrate moi- eties recognized suggest that serum lectins may contribute as a carbohydrate-based recognition system for potentially pathogenic microorganisms. One approach for the application of transgenic technology to disease resistance may be at the level of promoting the efficient recognition of the pathogen in the organism in ques- tion. Therefore, lectins may be considered as suitable target gene(s) for transgenesis since upon recognition of the pathogen. the complex lectin/pathogen would be phagocytosed and tngger the organism's natural defense mechanisms. In vitro manipulation of the lectin specificity aimed at the recognition of the appropriate pathogen target carbohydrate moiety, could be accomplished by ;ite-directed mutagenesis and the modified gene incorporated into the germline of the invertebrate of interest. Clearly, an exception to this approach would include those obligate or facultative intra- cellular pathogens that have developed strategies for blocking or evading the host's intracellular killing mechanisms. BIVALVE FISHERIES A PROFILE OF THE EAST HAMPTON TOWN SHELL- FISH HATCHERY AND RESEEDING PROGRAM. John Al- dred,* Town of East Hampton. 159 Pantigo Road. East Hampton. NY 11937. The East Hampton Town Shellfish Hatchery is located in Mon- tauk. the easternmost hamlet on New York's Long Island. Built in 1989-90 with seed money from New York State, it is operated by the Town for the enhancement of local shellfish stocks on public bottomlands. Ten percent of yearly production is made available to the state for regional distribution. The motivation for a facility of this kind was provided first by the closure of the local striped bass fishery in 1984 and in follow- ing years by the virtual elimination of the bay scallop from the region due to recurrent smothering algal blooms known as the brown tide. These two fisheries had historically furnished the larg- est earning potential for local inshore commercial fishermen. The hatchery annually produces in the neighborhood of ten million hard clam, oyster and bay scallop seed suitable for field distribution. A seven thousand square foot former U.S. Navy warehouse on Fort Pond Bay houses static water larval and pedi- veliger rearing systems, a flowing water upwelling nursery, tem- perature controlled as well as mass culture algal systems, shop, lab, and office space. A field nursery in Napeague Harbor, Ama- gansett contains rafted tray and pearl net systems for final grow out to planting size. A second upwelling nursery is located in Three Mile Harbor, East Hampton to take advantage of warmer harbor waters and to provide alternatives in the eventuality of unantici- pated water quality problems in one or another site. Along with Town officials, the hatchery has also organized oyster relays from uncertified waters to provide harvest potential and participated in designating management areas to protect spawning stocks. Projects underway include an overwinter sur- vival study of hard clam seed in different sediment types, a pilot oyster aquaculture project for local fishermen, a demonstration of bay scallop spat collection techniques for the Peconic estuary, and a comparative assessment of clam seed survival using hand and machine planting techniques. LEASE SITE CONSIDERATIONS FOR HARD CLAM AQUACULTURE IN FLORIDA. W. S. Arnold,* H. A. Nor- ris, and M. E. Berrigan, Florida Department of Environmental Protection. Florida Marine Research Institute. 100 Eighth Avenue S.E., St. Petersburg. FL 33701. Hard clams of the genus Mercenaria support an important com- mercial fishery in the Indian River lagoon on the eastern coast of central Florida. That fishery has supported landings with an esti- National Shellfisheries Association. Baltimore, Maryland Abslracl. 19% Annual Meeting. April 14-18. 1996 479 mated annual ex-vessel value of as much as $15 million. Unfor- tunately, because of the extreme variability in abundance of nat- ural hard clam stocks in the lagoon, the annual value of the fishery fluctuates drastically although market demand remains strong. Hard clam aquaculture provides a viable means of meeting market demand while avoiding the vagaries of natural clam sup- ply. The hard clam aquaculture industry is burgeoning in the In- dian River and is expanding statewide. It is more expedient to culture hard clams in Florida than it is in northern states because of the more rapid shell growth of Florida clams and the abundance of sites with suitable water quality in Florida. However, conflicts over resource allocation are arising between hard clam aquacul- turists and the harvesters of natural clam beds and between aqua- culturists and the resource managers concerned with the impact of aquaculture on the natural environment. The harvesters are con- cerned that lease sites, generally 5-10 acres in size, are being located in areas that support productive natural clam beds, thus denying harvesters access to those beds. Resource managers are concerned about the impact of aquaculture operations on the nat- ural benthic assemblage, including seagrass beds, and on the aes- thetics of the local environment. Potential lease sites are surveyed prior to letting, but that effort is conducted on a case-by-case basis, requires considerable effort by the lessee with no guarantee of success, and provides no coherent framework for lease alloca- tion. To mitigate these conflicts, we are developing Geographic In- formation System (GlS)-based tools to map the distribution of hard clam beds and other natural features. Information on the distribu- tion of clams and seagrass beds, water depth, proposed lease boundaries, physical features, and any other available and perti- nent information is fed into CIS map overlays. These map over- lays can then be plotted and presented to all concerned user groups for consideration. The maps can also be applied a priori to guide potential culturists in the selection of a suitable site, and to guide the State of Florida in the macroscale development of hard clam aquaculture in Florida. RELATIVE EFFICIENCY OF 3.5" DREDGE RINGS IN THE OFFSHORE SEA SCALLOP FISHERY. Jeffrey C. Brust,* William D. DuPaul. and James E. Kirkley, Virginia Institute of Marine Science. College of William and Mary, Gloucester Point VA 23062. The use of an average meat count restriction in the original sea scallop iPlacopeclen magellanicus) fishery management plan (SSFMP) was not effective in protecting small scallops from being harvested. In March 1994, Amendment #4 to the SSFMP was implemented. The major focus of the amendment was to decrease overall effort in the fishery. Age at first capture was to be con- trolled by a mandatory increase in the ring size used in the gear. Initially, ring size increased from 3.0" to 3.25" for 1994 and 1995. In 1996, the ring size increased to 3.5". This study used paired tows of the standard, 3.25" ring and experimental. 3.5" ring dredges on four commercial trips taken over a 10 month period during 1994 and 1995. The recruitment of a very large year class (1990) in late 1993 and early 1994 made it possible to assess the performance of the 3.5" ring dredge on a single year class as the scallops grew and more fully recruited to the gear. Our data indi- cate that the larger ring size will decrease the efficiency of the scallop dredge. This will effectively delay recruitment of an in- coming year class by as much as one year relative to the 3.25" ring dredge. The effects of this delayed entry have been evaluated relative to yield per recruit (YPR). spawning stock biomass (SSB), and age class structure of the resource. SHELLFISH MANAGEMENT PROGRAM— CLAM PRO- DUCTION. T. Jeffrey Davidson, Atlantic Veterinary College. University of P.E.I. . 550 University Avenue. Charlottetown, PEL CIA 4P3. Canada; Gef Flimlin, New Jersey Sea Grant Marine Advisory Service. 1623 Whitesville Road, Toms River, NJ 08755. In support of the hard shelled clam iMercenaria mercenaria) aquaculture industry in New Jersey, Shellfish Management Pro- gram— Clam Production has been developed. This computer- based program will enable aquaculture producers to more effec- tively manage their nursery and grow out plots in their leases. Parameters affecting productivity will be identified and evaluated allowing the producer to select those factors yielding the best economic benefits. Some of these parameters include: seed source and size, screen and bottom dynamics, water quality indicators (temperature, salinity, etc.) and relevant management factors. The inventory section will provide the producer with estimates of the number of animals in a plot, their movement within the lease, and mortality figures. Although Shellfish Management Program — Clam Production is a stand alone on farm program, data assimilated from a number of producers could assist in identifying and evaluating factors affect- ing the New Jersey clam aquaculture industry as a whole. LARVAL AND JUVENILE GROWTH OF STIMPSON'S SURFCLAM— A NEW CANDIDATE SPECIES FOR AQUA- CULTURE DEVELOPMENT? Christopher V. Davis,* Dar ling Marine Center. University of Maine. Walpole. ME 04573; Sandra E. Shumway, Southampton College. Long Island Uni- versity. Southampton. NY 11968. Stimpson's or Arctic surfclam (Maciroineris polynyma. Stimp- son 1860) is a cold water circumboreal species, distinguished from other surfclams by its purple colored foot, siphon and mantle edge which turn brilliant orange-red when cooked. Increased demand for wild surfclams to supply the Japanese sushi market prompted an investigation of the aquaculture potential for rearing this species in Maine waters. On four occasions over two years, adult clams were naturally conditioned and induced to spawn using temperature shock, sperm suspension and flowing seawater. Larvae were reared in either 40 480 Abslract. 1996 Annual Meeting, April 14-18, 1996 National Shellfisheries Association. Baltimore, Maryland or 400 liter conical tanks and fed daily a mixture of cultured microalgae. Prodissoconch 1 larvae developed in 24 or 96 hours depending on culture temperature (15.0 and 8.5°C respectively). Metamorphosis occurred in 24 to 42 days at 15.0 and 10.0°C respectively. Size of metamorphosing pediveligers varied from 320 ttm (SD = 22) shell length at I5.0°C to 271 |jim (SD = 26) shell length at 10.0°C. Juvenile growth was strongly influenced by substrate. Individuals reared on Nitex plastic screening m ambient sea water grew significantly slower than those reared in a silty/ sand sediment. Growth of juveniles occurred year round, varying from 0.009 mm ■ day" ' in the winter months to 0.102 mm • day" ' in late spring. Clams grew to 12.5 mm shell length, 0.29 g. live weight and 23.5 mm, 2.86 g. at one and two years respectively. After 29 months of growth, clams measured 31.4 mm in shell length and 4.93 g. live wt. These growth rates are approximately twice those seen in populations harvested from the w ild and may be further optimized through miproved culture meth- ods. This is the first report on the culture of Stimpson's surfclam beyond the larval stage. Given the rapid growth rates we observed, we believe this species may be amenable to aquaculture develop- ment in Maine, the Canadian Maritimes and the Pacific Northwest. A COMPARISON OF THE EFFECTS OF THREE DIFFER- ENT HABITAT MODIFICATIONS ON INTERTIDAL CLAM POPULATIONS IN PACIFIC NORTHWEST COASTAL ESTUARIES. Brett R. Dumbauld* and Martin Peoples, Washington State Department of Fish and Wildlife, P.O. Box 190, Ocean Park, WA 98640; David A. Armstrong, School of Fisheries. University of Washington, Seattle. WA 98195; Stephen G. Ratchford, North Carolina State University, Raleigh, NC 27650. A review of several field studies on the influence of epibenthic structural modifications to intertidal habitat in Pacific Northwest coastal estuaries suggests that several species of clams display similar responses to each. Three contemporary habitat modifica- tion issues are; invasion of the introduced cordgrass Spartina al- terniflora. the distribution of large quantities of oyster shell on intertidal tideflats to enhance the population of juvenile Dungeness crab. Cancer magister, and application of the pesticide carbaryl to remove thalassinid shrimp prior to oyster cultivation. Settlement measured as initial density oi Mya arenaria. Tapes japonica and Macoma spp. has been shown to be little affected by the physical structure of Spartina and shell with the exception of the large barrier presented by Spartina shoots in the later summer and fall. Subsequent survival of small clams however, is negatively influ- enced by the presence of epibenthic shell, which attracts predators such as juvenile Dungeness crab Cancer magister. Although tide height may preclude Dungeness crab from utilizing Spartina. sim- ilar declines in clam density were observed within cordgrass clones. After clams reach a size threshold, survival is less af- fected, but growth may be influenced depending on size of the clam, size of the structure, and hydrodynamic scaling factors. NATURAL AND EX-VESSEL MOISTURE CONTENT OF SEA SCALLOPS [PLACOPECTEN MAGELLAMCUS). William D. DuPaul,* Robert A. Fisher, and James E. Kirkley, Virginia Institute of Marine Science. College of William and Mary. Gloucester Point. VA 23062. Most of the sea scallop (Placopecten magellanicus) fishery is conducted on the continental shelf where scallops are shucked at sea. The adductor muscle, or scallop meats, are landed as the commercial product form. Scallop meats are generally stored in linen bags packed in ice for the duration of the fishing trip and are further processed at shore side facilities. Exposure of the scallop meats to ice melt or fresh water during storage or processing causes an increase in moisture content with concomitant increases in weight. During 1992, the U.S. Food and Drug Administration (FDA) became increasingly concerned with vessel handling and on-shore processing practices that substantially increased the moisture content of scallop meats. Subsequently, FDA ruled that a 'natural" scallop could not have a moisture content that exceeded 80'7f by weight. In 1995, Canada established a moisture standard not to exceed 81%. France has defined a moisture stan- dard for scallops based on a moisture/protein ratio not to exceed 4.99; 1. Prior to the present study, data on the natural moisture content of sea scallops in the northwest Atlantic Ocean has been limited and inadequate to establish regulatory provisions. The goal of this study was to assess natural and ex-vessel moisture content over an extended period of time (1990-1995) to account for latitudinal, depth, seasonal and individual variability. This more quantitative and verifiable study establishes statistically generated ranges of moisture contents and moisture/protein ratios and is discussed in view of input and output based regulatory strategies, international trade and implications for the industry. CHANGES IN HARD CLAM, MERCENARIA MERCE- NARIA, FISHERIES OF THE MID- ATLANTIC REGION IN RESPONSE TO STOCK FLUCTUATIONS. Gef Flimlin,* NJ Sea Grant Marine Advisory Service, 1623 Whitesville Rd., Toms River, NJ 08755. The processes by which the Hard Clam, or Northern Quahog. Mercenaria mercenaria, are being produced in the Mid-Atlantic region have changed dramatically in the past twenty years. Wild harvest in approved waters is often replaced by relay programs, shellfish depuration, and clam aquaculture. These activities are the response by the commercial industry to stock reductions in ap- proved water. Since participation by the commercial sector is linked to landings and harvest price, the number of clammers fluctuates also. Reasons for changes in relative abundance in coastal areas are speculative, but increased human populations, use of estuarine waters for cooling of electric generating plants, increased use of copper in dock building materials and antifouling paint, and the rise of outboard engine use are most often identified by the commercial industry as the most probable causes. Even in National Shellfisheries Association. Baltimore. Maryland Absinwi. 1996 Annual Meeting, April 14-18, 1996 481 areas of intense clam aquaculture, there are no significant exam- ples of new clam .sets. Since there arc no obvious reasons tor stock reductions, it is imperative that an ad hoc group of industry members, extension personnel and shellfish researchers be established to focus effort on identifying causes of stock reductions and ways to improve them, whether through stock enhancement, public aquaculture. or habitat improvement. POLYCULTURE OF SEA SCALLOP SUSPENDED FROM SALMON CAGES. Alexander Gryska, New England Fisheries Development Association. 4.'^1 D Street. Boston. MA 02210: G. Jay Parsons,* Aquatic Industries Ltd.. P.O. Box 294, St. An- drews, NB EGG 2X0; Sandra E. Shumway, Natural Science Division. Southampton College. L.I.U.. Southampton. NY 11968; Kristin Geib, P.O. Box 103. West Boothbay Harbor, ME 04575; Ian Emery, Snug Harbor Scallop Farm, P.O. Box 17200, Pembrooke, ME 04666; Sue Kuenstner. Ncu England Fisheries Development Association. 4-'^! D Street. Boston, MA 02210. Commercial culture of the sea scallop, Placopccten iinii;ellaii- icus. is an expanding industry in Atlantic Canada and New En- gland. In an experiment designed to examine the commercial fea- sibility of polyculturing scallops with Atlantic salmon, we mea- sured the growth and survival of sea scallops grown in suspension on two salmon aquaculture sites in northeastern Maine. One site was in Johnson Ccive. Passamaquoddy Bay and the other was located off Treats Island, near Lubec. Sea scallop spat ( 1 1 months of age and 10.2 mm shell height) were grown in standard pearl nets and were deployed on drop lines containing ten nets in August 1994. One drop line of ten nets was sampled about every four months and scallops were counted, measured for shell height, and tissues weights determined. Water samples for chlorophyll and scallop tissue samples for PSP phycotoxin content were also ob- tained. Scallop growth at the two sites was 35.2 and 38.5 mm shell height after four months and survival was >90'^. After one year, shell heights were about 49 and 57 mm. wet adductor muscle weights were 2.8 and 4.5 g, and growth rates were 0. 1 1 and 0. 13 mm per day. These growth rates were comparable to sea scallops cultured in Atlantic Canada. Reduced rates of survival were found during the latter part of the experiment and were attributable, in part, to heavy fouling by blue mussels. The potential for supple- mental income, diversification of the salmon aquaculture industry, and logistics of culturing scallop in conjunction with salmon will be discussed. THE 1995 STATUS OF THE SHELLFISHERIES FOR THE NORTHERN QUAHOG, MERCENARIA MERCENARIA (L.) IN NEW ENGLAND. Michael A. Rice, Dept. of Fishenes. Animal and Veterinary Science, University of Rhode Island, Kingston. RI 02881 . Fisheries for northern quahogs in the southern New England region have been in existence since pre-colonial times, and as recently as the middle 1980s the major fishery market source area. Since the late 1980s, there have been declinina catches in New England and increasing market supplies of quahogs from the Mid- dle Atlantic and Southern States due to relay, depuration, and aquaculture programs. Commercial quahog landings in the three major producing states in New England. Massachusetts, Connect- icut and Rhode Island were 34,659 bu (188 metric tons meat weight), 187,240 bu (1017 mt) and 134,417 bu (730 mt) respec- tively in 1994. Since 1990. landings of quahogs in Connecticut have increased largely due to introduction of containerized and bag relaying of product from conditional pollution closure areas. There has been a decline in the landings in Massachusetts and Rhode Island over the same period of time. The condition of quahog stocks in Massachusetts and Rhode Island are not particulariy poor, as catch per unit effort has been steady. Lowered catches in these states has been largely driven by the economics of the fish- ery. The low capitalization required by the quahog fishery (mostly bullrakes) allows fishermen to leave the profession as quickly as they can enter. For example during the mid-1980s when quahog prices were relatively attractive, there were about 800 full time shellfishermen in Rhode Island, but now there are only about 200. These remaining shellfishermen are individually catching as many quahogs as they had previously, but their overall income is down. It is recommended that greater attention to cooperative marketing by the fishermen can lead to greater economic returns. DEVELOPMENT OF THE HARD CLAM. MERCENARIA MERCENARIA, FISHERY IN THE SOUTH ATLANTIC RE- GION. Jack M. Whetstone,* Marine Extension Program, Clem- son University, P.O. Drawer 1 100. Georgetown, SC 29442-1 100; William D. Anderson, SCDNR, P.O. Box 12559, Charleston, SC 29442; Philip S. Kemp, Jr., UNC Sea Grant College Pro- gram, P.O. Box 3146, Atlantic Beach. NC 28512; Randal L. Walker, University of Georgia Marine Extension Service, 20 Ocean Science Circle, Savannah. GA 31411. The hard clam, Menenariu menenaria. fishery in the South Atlantic has a historical background to colonial times. Landings remained relatively low until the late 1970"s when the commercial fishery adopted mechanized harvesting, relaying and depuration techniques. The landings dramatically increased in the late 1970's but have stabilized since the early 1980's. Commercial landings for 1994 in the South Atlantic region were: North Carolina-607 metric tons (mt) valued at $6,756,631; South Carolina-133 mt valued at $1 ,068,260; and Georgia^. 9 mt valued at $73,158. Variation in landings between states can best be attributed to differences in gear and leasing regulations. Any increases in land- ings in the future will depend upon the development of hard clam aquaculture in the region, since the fishery appears to be at its maximum sustainable yield levels on open shellfish grounds. Since the late 1980's hard clam aquaculture has received increased interest in the South Atlantic region. Hard clam aquaculture has developed to a varying degree within each state due to the regu- lation of leases and seed importation requirements. 482 Abstract. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore. Maryland BIVALVE SEED PRODUCTION A COST-BENEFIT EVALUATION OF HATCHERY-PRO- DUCED OYSTERS IN MARYLAND. Mark L. Homer. Rob- ert Bussell, and Chris Judy, Maryland Department of Natural Resources, Piney Point Aquaculture Center. P.O. Box 150. Piney Point, MD 20674. In 1994. a pilot program was developed for the Maryland De- partment of Natural Resources' hatchery located at Pmey Point. MD. The goal of the program was to determine gross and net operating production costs and compare these with the State's Natural Seed Repletion Program. The hatchery production time- line began in March, with the acquisition and conditioning of brood stock, continued through spawning and larval development larval setting, and ended with the planting of two lots of spat on shell, one in September and the other in November. All costs. labor, materials, and utilities, were carefully tracked with a total of $37,500 spent on production, site preparation, and planting. Gross production cost, including site preparation and planting, was es- timated to be $0.0056 per spat. In September 1994. 3.4 million spat on shell {550 bushels) were transported to a 1 acre site, previously planted with 5,000 bushels of fresh shell, in Breton Bay (a tributary of the Potomac River). In mid-November 1994. 3.2 million spat on shell (350 bushels) were planted on a 2 acre portion of a natural oyster bar in the upper Wicomico River (a tributary of the Potomac). Monitor- ing of these oysters for growth and survivorship began immedi- ately after planting and has continued on a bimonthly basis. Initial mortality was extremely high, over 30%, attributed primarily to damage associated with transporting the oysters. As of September 1995, survivorship of the spat planted on the Breton Bay site stood at 8.6% and on the Wicomico River site 13.5%. The current net cost of the combined plantings stands at $0,049 per spat with these oysters needing about 9 more months to attain market size. The hatchery results were compared with cost and production estimates from the State's Natural Seed Repletion Program. Costs were estimated from experimental plots located near the mouth of the Chester River. Gross production cost was similar to the hatch- ery value, $0.0058 per spat, even though counts were unusually low for the natural seed that were moved to this site in November 1994. As of March 1995. survivorship exceeded 93% . giving a net production cost of $0.0063 per spat. This site will be surveyed again in November 1995. HATCHERY AND FIELD CULTURE TECHNIQUES FOR THE GIANT SEA SCALLOP PLACOPECTEN MAGELLANl- CUS. Richard C. Karney,* Martha's Vineyard Shellfish Group. Inc., Oak Bluffs. MA 02557; Frank A. Dutra, Nantucket Education and Research Foundation, Nantucket. MA 02554: David Dutra and ludy Dutra, Truro Aquaculture Project, North Truro, MA 02652. Under funding from the National Marine Fisheries Service. l-!:;iiing Industry Grants Program, the Martha's Vineyard Shellfish Grcjup adapted hatchery culture methods for bay scallops to the successful culture of the giant sea scallop. Field collected brood- stock were sufficiently ripe in early March (sea water temperature, 4-5°C) to spawn just over seven million eggs. The fertilized eggs were transferred to a 400 liter larval conical, with one micron filtered, aerated, sea water at 12C. After 48 hours, the conical was drained, and about three million scallops (ciliated blastulae and trochophores) were recovered and resuspended. Straight hinge lar- vae were not observed until the second drain down on Day 4. Larval culture protocol throughout the almost 40 day larval period included a daily feeding of Isochiisis galbana (T-ISO) and/or Cha- etoceros neogracili with a drain down and sizing every other day. The larvae were cultured in three 400 liter conicals of five micron bag filtered sea water, heated to about 15C (range I3-17C). Be- tween Days 28 and 38, 1,350,000 pediveliger larvae (about 250 microns) were moved to downweller sieves. The first fully set juvenile was observed on Day 32. Set scallops were cultured on downweller sieves ( 130-300 microns) with a flow of bag filtered water (10-50 microns) at sea water temperatures of 8-16C. By June 8 (Day 90) the largest seed measured 2 mm and were moved to 1 mm mesh Korean spat bags in a cage anchored off the shell- fish hatchery. By July 3 a total of 519.000 2 mm seed were successfully transferred to the inshore field culture systems. All of several potential off-shore growers were frustrated in their attempts to secure proper permits from the regulatory agen- cies. Over 80 percent of the 5 19.000 seed scallops were lost during the month of July when water temperatures reached 22C and reg- ulatory delays prevented transfer of the seed to cold water growout sites. At the end of July, the remaining 90.000 seed were finally permitted to be moved to the deep water (65') site of the Truro Aquaculture Project in Cape Cod Bay. The seed were stocked in 130 2.5 mm mesh pearl nets (600 seed/net) and ten 18" by 24" 'A" mesh plastic bags ( 1 .200 seed/bag), which were set into 75 tiers in five vinyl-coated, wire bottom cages. On September 29. after two months on the Cape Cod site. 66 percent of the seed scallops (X = 16 mm) were still alive and further thinned. The seed was sampled again on November 10 and averaged 22.4 mm. HOW IMPORTANT IS THE TIME OF TRANSPLANT IN THE SUCCESS OF OYSTER (RELAY) FARMING? Eric Powell* and Susan Ford. Haskin Shellfish Lab., Rutgers Univ., Port Norris. NJ 08349; John Klinch and Eileen Hofmann, CCPO. Old Dominion Univ.. Norfolk. VA 23529. The oyster industry in New Jersey transplants oysters from seed beds to leases and then later harvests them for sale. During their time on the leased grounds, oysters grow to market-size, but also suffer increased mortality from disease. At one time, oysters were left on the leased grounds for more than one year. Recently, losses to disease have forced the industry to rely on oysters kept on leases for only a few months. Normally, transplant occurs in May /early June and harvest from September to December. Poor survival over the summer led to the recommendation that transplanting occur one month earlier, in April, to permit harvest before disease mor- National Shellfisheries Association. Baltimore. Maryland Ahstracl. 1996 Annual Meeting, April 14-18, 1996 483 tality becomes significant. Results of a one year trial by the in- dustry suggest an improved fall harvest. The earlier transplant was likely successful because the spring bloom is stronger over the leased grounds than over the seed beds. The degree to which the spnng bloom can be advantageously used may be crucial. We used an oyster population dynamics model to investigate the timing of transplant. Simulations were run for transplants in November. January, March, April, and May. The simulations agree with observation in suggesting that an increased harvest results from transplanting oysters in April rather than May/June. The simulations also indicate that earlier transplants are even more advantageous. A November transplant nearly doubles the available harvest. A March transplant is only moderately less advantageous than November. In both cases oysters take advantage of the ma- jority of the spring bloom and this increased growth in the spring increases the abundance of market-size oysters in the spring and fall. Mortality rates from P. marimis decline, but not dramatically. Thus the simulations suggest that the principle effect of a change in transplanting time is to change the abundance of market-size oysters prior to the initiation of disease mortality in July, and, thus, at the same mortality rate, more oysters are available for harvest in the fall. A POSSIBLE OPTION FOR ENHANCEMENT OF THE WILD FISHERY FOR THE SEA SCALLOP, PLA- COPECTEN MAGELLAMCUS. Shawn M. C. Robinson,* Jim D. Martin, and Ross A. Chandler, St. Andrews Biological Station, Dept. Fisheries and Oceans. St. Andrews, New Bruns- wick, Canada, EOG 2X0; Don Bishop, Fukui North America, P.O. Box 1 19, Island View Drive, Golden Lake. Ontario, Canada, KOJ 1X0. To date the wild scallop fishery in eastern Canada has relied on the natural cycles of recruitment in the various beds of sea scallops to sustain their fishing activity. As with other populations of scal- lops around the world, the sea scallop is subject to variable re- cruitment rates which are most likely due to biological and phys- ical factors in the larval and early juvenile stages. With increasing fishing pressure on the scallop stocks, the industry is beginning to look at methods of stabilising and enhancing their production. The goal of our research was to test the feasibility of enhancing the recruitment rate of scallops to the bottom by collecting scallop larvae using a suitable substrate for settlement which would also allow the resulting spat to detach and settle to the bottom at a later date. We deployed replicates of five different types of Biocord (re- sembling a fuzzy type of rope made for biological filters) in Sep- tember 1995 using divers and monitored the settlement and sur- vival of the early juveniles until early December. Results indicated all five substrates were attractive as a settlement substrate although there were differences between the different types. Settlement den- sities ranged from 100 to 400 spat/m of biocord. Two waves of settlement were also observed. SELECTIVITY AND ECONOMIC ANALYSIS OF DIFFER- ENT CLLTCH MATERIALS FOR OYSTER SETTING IN HACKBERRY BAY, LA. Mark Schexnayder, Randall Pausina, Ron Dugas, and David Lavergne. Setting rates of eastern oyster, Crassosirea virginica (Gmelin, 1791). on six kinds of cultch. were compared from data collected on the "Public Seed Ground" in Hackberry Bay and from a tray experiment in southern Barataria Bay, LA. Crushed concrete, shucked oyster shell, reef shell, mixed shell, Kentucky limestone, and Bahamian limestone were placed in large volumes in Hack- berry Bay during August 1994. Thus producing an actual applied experimental situation. Square meter samples of these planted ma- terials were taken in November 1994 and July 1995. On an equal volume basis, crushed concrete produced the most seed oysters; shucked shell production was a close second. In the trays, crushed concrete again produced the most seed oysters per volume. Cal- culation of cost per sack of seed oysters for each of the six mate- rials indicated that crushed concrete, shucked oyster shell, and mixed shell were the most economical, while Bahamian and Ken- tucky limestones were the least. On silty-clay bottoms shucked oyster shell as cultch offers several advantages over crushed con- crete, including more exposed surface for setting and less density. hence slower sinking rates. CONSERVATION OF FRESHWATER MUSSELS FACTORS INFLUENCING THE GROWTH AND SUR- VIVAL OF JUVENILE VILLOSA IRIS (BIVALVIA: UNION- IDAE) IN AN ARTIFICIAL STREAM SYSTEM. Braven B. Beaty* and Richard J. Neves, Virginia Cooperative Research Unit. Department of Fisheries and Wildlife Sciences. Virginia Polytechnic Institute and State University. Blacksburg. VA 24061. The propagation of freshwater mussels is being promoted to assist in the recovery of declining wild populations. This project investigated some of the properties that influence the success of a juvenile mussel rearing system. Newly transformed juvenile mus- sels (Villosa iris) were placed in a flow-through artificial stream system supplied with natural river water from the Clinch River in Carbo, VA. Juvenile mussels were held in the troughs in small containers loaded to one of two fixed depths with substrate of two size fractions; less than 120 |j.m and between 120 (xm and 600 p.m. Individual containers of juveniles were removed at intervals to assess the survival and growth of the animals. Mean survival rates were 40.8%, 17.3% and 19.1% at days 30. 74 and 98, respec- tively. Size, as approximate area of the valve, was 0.839 mm". 1.496 irmi" and 1.313 mm" at days 30. 74 and 98. respectively. Little growth or mortality occurred after the day 74 sampling in- terval in mid-October. Below about 14°C. the juveniles ceased 484 Abstract. 1996 Annual Meeting, April 14-18. 1996 National Shellfisheries Association. Baltimore, Mar>'iand growing and suffered little mortality. The effect of flow on sur- vival and growth was determined at each time increment. Flow had no effect on survival after 30 days or on growth at any time, but a significant effect on survival after 74 and 98 days. Increased flow caused a decrease in the survival rate after 74 and 90 days {R' of 31.9% and 27.2%, respectively). Initial substrate size and depth had no effect on either survival or growth at any sampling time. This is likely because a layer of fine silt settled from the overlying water upon the substrate, and 75-94% of the mussels were found in this fine silt. Research results could be used to design more effective rearing systems for juvenile mussels. DECLINE AND DECIMATION: THE EXTIRPATION OF THE UNIONID FRESHWATER BIVALVES OF NORTH AMERICA. Arthur E. Bogan, Freshwater Molluscan Research. 36 Venus Way, Sewcll, NJ 08080. North America north of Mexico historically was home to the most diverse freshwater bivalve fauna in the world with approxi- mately 300 species. The first real attempt to assess the status of freshwater bivalves began in 1970 and by 1971, II freshwater bivalve taxa were presumed extinct and generally, 120 taxa were considered rare or endangered. In 1973. the Endangered Species Act was passed and the U.S. Fish and Wildlife Service began evaluating the status of species and listing some as threatened or endangered. As a result of increased interest and the federal listing of species, various states began listing freshwater bivalves as lo- cally extirpated, threatened or endangered. The American Fisher- ies Society, in the Common and Scientific Names of Aquatic Invertebrates from the United States and Canada: Mollusks (1988), listed 13 taxa as extinct and 30 taxa as federally endan- gered. Today at least 35 freshwater bivalve taxa are presumed extinct, 52 taxa endangered. 5 taxa threatened at the federal level and an additional 70 taxa as candidates for either threatened or endangered status. At this time 12% of this fauna is presumed extinct, 42% is listed or to be listed and an additional 25% of the fauna is declining. Less than 25% of the freshwater bivalve taxa appear to be maintaining stable populations today. The problems with the fouling and pollution of the freshwaters of North America were recognized early. S. N. Rhoads (1895) documented the decimation of the freshwater bivalve fauna in the lower Monongahela River. A. E. Ortmann reported the decima- tion of the unionid fauna in western Pennsylvania ( 1909) and in the Pigeon River in East Tennessee (1918). The central problem lead- ing to the decline and decimation of the freshwater unionid fauna is the modification and destruction of their aquatic habitat with sedimentation as the single major factor. Sources of sedimentation include poor agricultural and timbering practices. Damming of major rivers has had a dramatic impact on this fauna with the loss of obligate host fish due to changes in water quality and loss of habitat. In-stream gravel mining, dredging, channelization and the often associated headcutting has eliminated stable mussel habitat. Acid mine drainage, and various point and non-point pollution sources all continue to decimate local unionid populations. A new threat to the continued survival of unionid taxa is the introduction m the mid-1980's of the zebra mussel (Dreisseiia polytnorpha) . These small, bysally attached bivalves cover and smother the na- tive mussels. GEOGRAPHIC VARIATION IN UNIONID GENETIC STRUCTURE: DO MANAGEMENT UNITS EXIST? David J. Berg,* Dept. of Zoology, Miami University, Hamilton, OH 45011; Sheldon I. Guttman, Dept. of Zoology, Miami Univer- sity, Oxford, OH 45066; Emily G. Cantonwine, Savannah River Ecology Laboratory, Aiken, SC 29802. Unionid populations contain significant levels of within popu- lation (w-p) genetic variation. Little is known about levels of among population (a-p) variation and the degree of difference among geographically separated populations. We used allozyme electrophoresis to: 1 ) describe w-p variation for populations of Quadniki quadnda from the Ohio (OH), Tennessee (TN), and Mississippi (MI) river basins; 2) compare a-p variation among basins; 3) consider the results in the context of freshwater mussel conservation biology. All populations contained significant w-p variation ( 1 .8-2.6 alleles/locus; >62% polymorphic loci, mean H = 0.25 to 0.32). OH and TN did not contain significant levels of a-p variation, but a population from the smaller Tensas River, LA (TR) had different allele frequencies at several loci. Genetic dis- tance was positively correlated with geographic distance (r^ = 0.721, p < O.OOI, n = 19); TR was distinct from the other populations. We examined these results using the concept of Man- agement Units (MU) — populations that are functionally indepen- dent and exhibit significant levels of mitochondrial or nuclear a-p variation. In the MS basin, at least 2 MUs are present. These may be based on geographic distance or river size. Successful conser- vation of unionids requires that issues of genetic structure be con- sidered when developing management plans for threatened and endangered taxa. CONTAMINANT IMPACTS ON NATIVE FRESHWATER MUSSELS— LETHAL AND SUBLETHAL RESPONSES RELATIVE TO WATER QUALITY CRITERIA. Anne E. Keller, National Biological Service, Gainesville, FL 32653. Native freshwater mussels (Family; Unionidae), the most imperilled fauna in North America, attain their greatest diversity in the southeastern United States. These sedentary animals oc- cupy a unique niche in aquatic systems because they have a par- asitic stage during early development, become free-living filter feeders, dwell at the water/sediment interface, and live for up to 90 years. Many envu-onmental impacts have been identified as con- tributors to the loss of mussel fauna, including habitat destruction, loss of host fish and competition with nonindigenous mollusks. However, contaminants are believed to present one of the most serious challenges to the continued survival of many unionid spe- cies. National Shellfisheries Association, Baltimore. Maryland Ahslracl. 1996 Annual Meeting. April 14-18. 1996 485 While unionids have been used as indicators of environmental quality, little was known about their sensitivity to contaminants until the last five to ten years. Most information currently available is on acute toxicity. Little is known about the effects of long term (chronic) exposure. Their decline relative to contaminant toxicity needs examination because the input of pollutants into our aquatic systems can be regulated. Water quality criteria, risk assessments, and current research on sublethal impacts of contamination on unionids will be discussed. In addition, a review of available EC50 and LC50 data will be provided. The utility of such information in addressing conservation issues will be emphasized. MALATHION TOXICITY TO THREE LIFE STAGES OF UNIONID MUSSELS. Anne E. Keller* and Shane Ruessler. National Biological Service. 7920 NW 71st St.. Gainesville. FL 32653. The acute toxicity of malathion to clochidia. juveniles or adults of seven species of freshwater mussels was determined in soft or moderately hard reconstituted test water at either 25°C or 32°C. Glochidia tests were conducted for 4. 24 or 48 h. while juvenile (cultured by in vivo and ;/; viiro methods) and adult exposures lasted 96 h. LC50 values for glochidia of Uncrbackia imbecilis and L. teres tested at 25°C in soft water were 447 mg/L (48 h) and 28 mg/L (4 h) respectively, while Villosa Uenosa glochidia had an LC50 or 1 19 mg/L (48 h) at 32°C in moderately hard reconstituted water. Tests with juvenile mussels produced 96 h LC50s ranging from 40 mg/L for U . imbecilis in 32°C soft reconstituted water to 219 mg/L at 25°C in moderately hard water. No LC50s were calculable for adult mussels of three species in concentrations up to 350 mg/L. These values are considerably higher than the 48 h LC50 of 1 |J.g/L for D. magna. THE IMPORTANCE OF HABITAT HYDRAULICS IN THE RESTORATION OF NATIVE FRESHWATER MUSSELS. James B. Layzer, National Biological Service. Tennessee Coop- erative Fishery Research Unit, Tennessee Tech. Univ.. P.O. Box 5114. Cookeville. TN 38505. Freshwater mussel populations in North America have been devastated by a wide array of physical and chemical perturbations. In some cases, habitat destruction and the loss of mussel popula- tions is essentially permanent as in the case of the construction of dams which inundate riverine habitat, change water quality, and eliminate hosts fish populations. In many other cases, the factors responsible for the extirpation of mussel populations have largely been corrected and conditions may now be suitable for the rees- tablishment of mussels: however, it is suggested that during the intervening time between the extirpation of mussels and improve- ment in stream conditions other factors affecting stream hydraulics may prevent the successful reintroduction of mussels. In particu- lar, land-use practices within watersheds may have profoundly affected stream hydrographs by increasing peak discharges follow- ing precipitation and decreasing base flows during dry periods. Lower base flows may expose mussel beds, eliminate settlement of juveniles from otherwise suitable habitat, and affect host fish population dynamics and movements. Conversely, results of re- cent research indicate that high shear stress associated with peak discharge is likely responsible for unsuccessful settlement of ju- venile mussels in a headwater stream. Measuring or modelling simple hydraulic variables such as mean water column velocity is inadequate for assessing the affects of altered stream hydrographs on potential mussel habitat. In contrast, complex hydraulic vari- ables such as shear stress and Reynolds boundary number are potentially better predictors of hydraulically suitable sites for mus- sel reintroductions. DELAYED REPRODUCTION OF THE FRESHWATER MUSSEL ELLIPTIO COMPIJlNATA THROUGH TEMPER- ATURE AND PHOTOPERIOD CONTROL. William A. Lel- lis* and Connie S. Johnson, National Biological Service. Wells- boro. PA 16901. Elliptio complanala is a species of freshwater mussel com- mon to streams and rivers of the Atlantic slope. Egg fertiliza- tion, larval brooding, and glochidial release are reported to oc- cur within a period of several weeks during early- to mid-sum- mer. In this study we tested the ability of photoperiod and water temperature manipulation to prolong the availability of glochidia for use in artificial propagation. Brood mussels were collected from Pine Creek. Tioga County. PA in late December 1994 when water temperature (0.5°C) and photoperiod (9L;I5D) were sea- sonally low. Mussels were housed in groups of 46 within eight 1.2-m diameter circular fiberglass tanks containing 25 cm of gravel substrate and subjected to one of four environmental treat- ments. In the first treatment, photoperiod and temperature matched natural conditions. In the second and third treatments, winter conditions were prolonged for periods of 6 and 12 weeks beginning January 1. The fourth treatment matched natural conditions except that winter temperature was held constant at 10°C. Mussels subjected to natural conditions released glochidia between 16 and 19 C while photoperiod and temperature were rising. Initial conglutinates were white and leaf-shaped containing mostly immature glochidia. Subsequent conglutinates were com- posed of fully developed glochidia packaged within a clear matrix, while final release occurred as free individuals extruded on thin mucus-like strands. Prolonged winter conditions delayed repro- duction proportional to length of treatment, whereas elevated win- ter temperature had no effect on timing of glochidial release. Data indicate that the seasonal availability of Elliptio complanala glochidia can be extended three-fold using photoperiod and tem- perature manipulation. 486 Ahsiracl. 1996 Annual Meeting. April 18. 1996 National Shellfisheries Association. Baltimore, Maryland FRESHWATER MUSSEL CONSERVATION AND THE EN- DANGERED SPECIES ACT. Debbie C. Mignogno, US Fish and Wildlife Service. 300 Westgate Center Drive. Hadley, MA 01035-9589. The repercussions of nearly unprecedented losses in biodiver- sity are explicitly depicted in North American freshwater mussel fauna. The continent exhibits the richest array of freshwater mus- sel fauna in the world and 70% of recognized species are consid- ered endangered, threatened, or of special concern. The authorities conferred by the Federal Endangered Species Act (ESA) and var- ious state inacted Endangered Species Acts mandate the develop- ment of programs for the conservation of listed and candidate freshwater mussel species by the U.S. Fish and Wildlife Service and other Federal and State management agencies. 1 will examine various components of the ESA and outline recent research, pres- ervation, and recovery activities conducted by Federal and State agencies under the auspices of the ESA for freshwater mussels. with particular emphasis on the northeastern United States. In addition. I will discuss threats to the continuation of these pro- grams from the Congressionally-directed budget cuts and morato- riums placed on activities conducted pursuant to the ESA. THE EXOTIC ZEBRA MUSSEL IN NORTH AMERICA: A DIRE PROGNOSIS FOR NATIVE FRESHWATER MUS- SELS (UNIONIDAE). Richard J. Neves, National Biological Service. Virginia Cooperative Fish and Wildlife Research Unit. Virginia Tech, Blacksburg. VA 24061; Catherine Gatenby and Bruce Parker, Biology Department. Virginia Tech. Blacksburg, VA 24061. The zebra mussel (Dreissena polymorphu) is running rampant in North American waters, infesting and exterminating native unionids from much of Lake St. Clair, western Lake Erie. Detroit River. St. Lawrence River, and other localized areas. Its escape in 1991 from Lake Michigan to the Illinois River and spread through- out the Mississippi River Bason now jeopardizes the native mussel fauna in many major tributaries. Populations expand rapidly and achieve densities of greater than 50.000/m" in 2 years of coloni- zation. The species readily colonizes living unionids by byssal threads, with densities exceeding 10,000 individuals and weights up to four times that of the colonized unionid. Negative effects on unionid populations include inhibition to feeding and respiration. reduction in glycogen reserves, and ultimately death within 5 years of infestation. To anticipate possible extirpations or extinctions of native mus- sel species, a project was initiated in 1992 to evaluate the feasi- bility of using small ponds and headwater rivers as refugia for unionids in the Ohio and Tennessee rivers. Most unionid species confined in cages and pocket nets in ponds from 1992-1995 have •■xhibited good survival (>70'7f). and studies are evaluating re- ]>:i-ductive success in ponds and a river refugium. A geographic ne.vork of natural and artificial refugia may be necessary to pre- vent an unprecedented spasm of extinctions of native riverine unionids. CRUSTACEAN BIOLOGY AND FISHERIES SOME RECENT TRENDS IN MARYLAND BLUE CRAB POPULATIONS. George R Abbe,* Estuarine Research Center, Academy of Natural Sciences. 10545 Mackall Road, St. Leonard, MD 20685; Cluney Stagg, Fisheries Administration. Maryland Department of Natural Resources. Tawes B-2. 580 Taylor Ave- nue. Annapolis. MD 21401. With major reductions in the size of many Maryland Chesa- peake Bay fisheries, added pressure has been exerted on the blue crab in recent years. In an effort to understand some of the con- sequences of this increased pressure, we have analyzed data col- lected along 12.5 km of western Chesapeake Bay in Calvert County from 1968 to 1995. Commercial peeler crab pots of 25- mm ( 1-in) mesh, baited daily with menhaden, were used to sample crab stocks at three locations with up to sixty pots fished during alternate weeks from June through November. Station catches were sorted, measured and weighed by sex. From 1968 through 1995 112,994 crabs were caught in 18.106 pots, of which 73% were legal size. Although the annual mean catch per unit effort (CPUE) showed considerable variation, this appeared to be nor- mal. Crabs per pot ranged from 0.85 in 1968 to 20.01 in 1981 while Maryland commercial landings ranged from 10.3 million pounds to 59.7 million pounds during the same years. From 1968 to 1980 legal CPUE averaged 3.60. from 1981 to 1985 it averaged 8. 14. and from 1986 to 1995 it was 3.66. Thus the legal CPUE of the most recent period is little different from that of the earliest period. There are. however, several trends that have become ap- parent in recent years indicating that fishing pressure may be put- ting a severe strain on the blue crab population. Significant cor- relation between this fishery independent data and Maryland DNR's fishery dependent data demonstrate the relevance of these trends. From 1968 to 1982 the annual male percentage decreased sig- nificantly from 66% to 38% (r" = 0.79; p < 0.01). Since 1983 this percentage has fluctuated more, but it has not shown any further decrease. Mean carapace width and weight of females have not changed significantly over time, but width of males (r^ = 0.47) and weight of males (r" = 0.34) have both decreased sig- nificantly (p < 0.01 ). Legal size, which constituted 64 to 86% of the annual catch between 1968 and 1991. has had its three lowest years since just 1992; and the percentage of legal males in the catch decreased from 56% in 1968 to 19%> in 1995 (13% in 1994) (r" =0.75; p < 0.01). These most recent downward trends related to size of males indicate that they are being removed from the population shortly after reaching legal size. With only small males available to crabbers, even more pressure may exerted on females which could result in continued decreases in population size and stability. Further regulations may be necessary. National Shellfisheries Association. Baltimore. Maryland Abslracl. 1996 Annual Meeting. April 14-18, 1996 487 THE GONOPOD TEGUMENTAL GLANDS OF SNOW CRAB, CHIONOECETES OPILIO: A CLOSER LOOK YIELDS EVIDENCE FOR SEXUAL FUNCTION. Peter G. Beninger* and Annie Ferguson, Departemeni de Biologic. Uni- versite de Moncton. Moncton N.B. Canada ElA 3E9: Carole Lanteigne, Centre Marin de Shippagan. C.P. 1010. Shippagan. N.B. Canada EOB 2B0. The role of the tegumental glands found m the first gonopod of brachyuran crabs has hitherto been a matter of conjecture. In order to elucidate the nature and ultimate function of these glands, his- tological and histochemical studies were performed on 17 male snow crabs. Mature (M) and immature (IM) individuals were dif- ferentiated based on the carapace width (CW); cheliped height ratio. Immature crabs were subdivided into 3 groups: small im- mature (<40 mm CW), medium immature (40-70 mm CW), and large immature (70-100 mm CW). The precise distribution of the glands within the first gonopod was determined via serial sections (7-10 jj.m), and histochemical tests were performed for lipids, aminated substances, acid mucopolysaccharides, and neutral mu- copolysaccharides. The volume fraction of the gonopod glandular region occupied by glands was assessed using stereologic counts. The glands were determined to be of the rosette type, and restricted to a specific region at the base of the endopodite. Ducts leading from these glands to the cuticle of the ejaculatory canal only were clearly visible in medium immature to mature individ- uals; these ducts connected to pores in the cuticle. Cuticular pores and ducts were not observed in small immature crabs. The volume fraction of the glands increased in each successive maturity cate- gory, with a mean of 0.8% in small immature crabs and a mean of 8% in mature crabs. The glands contained either acid or neutral mucopolysaccharides, or a mixture of both. The pores of the ejac- ulatory canal contained similar secretions. These observations sup- port the conclusion that the first gonopod tegumental glands in C. opilio are accessory sex glands. THE EFFECT OF WATER VOLUME AND SURFACE AREA OF CULTURE CONTAINERS ON WEIGHT GAIN OF JUVENILE FRESHWATER PRAWNS, MACROBRACH- lUM ROSENBERGl. Louis R. DAbramo,* Curtis G. Sum- merlin, William H. Daniels, and H. J. Wan. Department of Wildlife and Fisheries, Mississippi State University, Mississippi State, MS 39762. Juvenile freshwater prawns (mean weight = 0.225 g) were individually held in containers under conditions of different vol- ume, surface area, and flow rate (turnover rates of 44 and 86 minutes), (2x2x2 factorial design) and fed a semi-purified diet. Reduced volume or reduced area was associated with a significant reduction in weight gain after 60 days. No significant effect of turnover rates was observed. The density effect on growth rate occurs even in the absence of other conspecifics and may be due to the attainment of a threshold level of a particular metabolite. The freshwater prawn exhibits the same density dependent growth relationship evidenced by other aquatic organisms. FIELD EXPERIMENTS ON THALASSINID SHRIMP CON- TROL FOR OYSTER CULTURE IN WASHINGTON STATE. Brett R. Dumbauld,* Washington State Department of Fish and Wildlife, P.O. Box 190, Ocean Park, WA 98640; David A. Armstrong and Kristine L. Feldman, School of Fisheries, University of Washington, Seattle, WA 98195; John R. Skalski, Center for Quantitative Sciences, University of Washington, Se- attle, WA 98195. Field experiments are being conducted to examine use of the pesticide carbaryl and the placement of oyster shell as a physical barrier to control mud shrimp Upogebia pugettensis and ghost shrimp Neonypaea californiensis on intertidal oyster culture grounds in Washington State coastal estuaries. Survival and growth of juvenile oysters Crassostrea gigas and re-invasion of shrimp were monitored on 100 m~ carbaryl treated plots and un- treated controls from 1990-1993. Addition of a thick layer of oyster shell to carbaryl treated and untreated mudflat is currently being investigated on 64 m' plots. Results of these expenments suggest marked differences between the effects of each species of shrimp on oyster seed and epibenthic shell. Ghost shrimp cause immediate loss of oyster seed as well as the oyster shell barrier, while mud shnmp have a lesser effect. All shrimp are killed when carbaryl is applied in July including small mud shrimp which settle as post-larval recruits in May and June. Ghost shrimp can re- invade treated plots almost immediately as post-larval recruits set- tling in fall and preferentially select open mud, while mud shrimp re-invade the following spring and appear to preferentially seek shell. Despite obvious deleterious effects of shrimp on oyster seed survival, no significant effects have been detected on oyster growth. RELATIONS AMONG FIXED STATION BLUE CRAB POT SAMPLING RESULTS, REPORTED CHESAPEAKE BAY LANDINGS AND WINTER DREDGE SURVEY RESULTS. Cluney Stagg,* Fisheries Administration, Maryland Department of Natural Resources, Tawes B-2, 580 Taylor Avenue, Annapolis, MD 21401; George R. Abbe, Estuarine Research Center, Acad- emy of Natural Sciences, 10545 Mackall Road, St. Leonard, MD 20685. Declining trends in available measures of apparent abundance have motivated a renewed interest in the status of the blue crab stock in Chesapeake Bay. One of the longest, and potentially most useful, data sets available for assessing the current status of the Chesapeake Bay blue crab stock was begun in 1968 and has continued to the present. Catch per unit effort (CPUE) and biological data (size and sex composition) were collected in com- mercial peeler pots of 25-mm mesh at three stations along 12.5 km of Chesapeake Bay in Calvert County. Correlation analysis was conducted to examine the relation between Calvert Cliffs 488 Abstract. 1996 Annual Meeting, April 14-18, 1996 National Shellfisheries Association, Baltimore, Maryland CPUE and reported (1) Maryland pot catch, (2) Maryland total catch, (3) Chesapeake Bay pot catch, and (4) Chesapeake Bay total catch. From 1968 to 1981, correlation coefficients (r) of (1 ) 0.91*, (2) 0.94*, (3) 0.82*, and (4) 0.82* resulted (* = p < O.OOI). By the end of the 1994 season, rs had declined by as much as 25%. It was hypothesized that the post- 1981 divergent trend could be largely explained by the sharp increase in crab pot effort in Maryland having compensated for changing blue crab abun- dances, such that catch did not move in parallel with Calvert Cliffs CPUE. Maryland reported commercial crab pot effort increased by about 50% from 1984 (first year available) through 1994. Follow- ing effort-standardization of Maryland reported crab pot catches from 1985 to 1994, there were improvements in r of as much as 18%. In an unrelated study supported by the National Marine Fish- eries Services's Chesapeake Bay Stock Assessment Committee, a randomized winter dredge blue crab survey has been conducted over all of Chesapeake Bay since the winter of 1988-89. More than 1000 samples have been taken each winter while crabs are dormant. A close relation between the dredge survey data and the Calvert Cliffs data would further support the representative char- acter of the Calvert Cliffs data. To assess the relation between the two, Calvert Cliffs annual CPUE was regressed against the prior winter nominal age-1 crab index (60-120 mm), with the following result: dredge survey age-l's predicted the Calvert Cliffs abun- dance index for the following summer with a coefficient of deter- mination, r", = 0.90 (p < 0.0001). The results from these two analyses imply that Calvert Cliffs data is representative of the bay wide blue crab stock, and thus, is a very useful long-term times series for evaluating the Chesapeake Bay blue crab stock status. Given the previous statement, Calvert Cliffs CPUE results indicate that current (1990-94) abundance levels more closely resemble those of the late- 1960s and 1970s, and that eady-to-mid 1980s abundances were probably anomalous. CRUSTACEAN HEALTH PROBLEMS IMPACT OF "BUMPER CAR" DISEASE ON THE NORTH AMERICAN LOBSTER FISHERY. Richard J. Cawthorn, Atlantic Veterinary College, University of Prince Edward Island, Chariottetown, P.E.I. CIA 4P3. Although 1993 landings of lobsters were valued at $300 million in Canada and $210 million in the United States, postharvest losses are 10-15% annually. ■"Bumper car" disease of lobsters caused by the scuticociliate Anophtyoides haemophila. can be sig- nificant in coldwater impoundments. Although outbreaks occur more frequently and with greater severity, the epidemiology and economic impact of "bumper car" disease are not well docu- mented. The ciliates are maintained in cell-free, defined medium at 5°C. Cultured ciliates require longer and more parasites to kill lobsters than those transmitted by intrahaemocoelic injection from lobster to lobster. Horizontal transmission likely occurs across gills of lobsters. Several licensed disinfectants and chemothera- peutants are efficacious against A. haemophila in vUro. Additional to indirect fluorescent antibody testing utilizing monoclonal anti- bodies prepared to sonicated ciliates, parasites are detected with oligonucleotide probes based on ssu-rDNA of A. haemophila. The prevalence of A. haemophila in wild-caught lobsters should be reevaluated with more sensitive and specific diagnostic tools. A definition of healthy versus ciliate-infected lobsters is being pre- pared, based on hacmatology and clinical chemistry of haemo- lymph. Our novel bar-coded labelling system for aquatic organ- isms facilitates experimental design and randomization protocols of lobsters. The model of "bumper car" disease will aid study of health and infectious disease processes of lobsters. (Funded as subcontracts from the Canadian Atlantic Lobster Promotion Asso- ciation (CALPA) and Diagnostic Chemicals Limited (DCL). CALPA and DCL were supported m part by the Industrial Re- search Assistance Program of the National Research Council of Canada.) AN OVERVIEW OF PENAEID SHRIMP PATHOGENS IN U.S. WATERS. John A. Couch, Sr. Res. Scientist, Gulf Ecol- ogy Division Laboratory, U.S. Environmental Protection Agency, 1 Sabine Island Dr.. Gulf Breeze. FL 32561. Intensive efforts, over the last 20 years, to culture penaeid shrimp in North America and worldwide have increased the focus on pathogens that restrict or limit success in shrimp survival. Since the first discovery and description of a pathogenic baculovirus in 1970-71 {Baculovirus penaei-^?) in penaeid shrimps, many other BP-type baculoviruses have been reported from at least six geo- graphic areas as pathogens of over 15 species of penaeid shrimps in 5 of the six subgenera of the genus Penaeus and in other related pcnaeioid genera (Lightner, 1993. and unpubl. Data). High larval shrimp mortalities are reported frequently caused by these and other baculoviruses in shrimp hatcheries. Apart from the dynamic role of newly discovered viruses in penaeids, other pathogens including bacterial species have been found to cause severe losses. Texas necrotizing hepatopancreatitis (a rickettsial form) and Vibriosis epizootics in North and South America are recently more clarified threats, with a longer history. These newly recognized pathogens are silhouetted against the background of known fungal, and protozoan pathogens that have been encountered for decades in penaeid shrimp in North America. These include, particularly, the microsporidian protozoan and sus- pect ciliate ecto-commensals whose specific roles are still unre- solved in shrimp/crustacean health. Future direction in research on viral, bacterial and protozoan pathogen of shrimp, particularly molecular probe use for diagnostics are noted. National Shcllfisheries Association. Baltimore. Maryland Abstract. 19% Annual Meeting. April 14-18. 1996 489 TOXICANT EFFECTS ON GRASS SHRIMP EMBRYOS. William S. Fisher* and Lee A. Courtney. National Research Council, Environmental Protection Agency. Gulf Breeze, FL 32561; Patricia S. Glas and James R Rayburn, Gulf Ecology Division, National Health and Environmental Effects Research Laboratory, Environmental Protection Agency. Gulf Breeze. FL 3256 L Embryos of the grass shrimp Paluemoiu'lcs piigio have been used successfully to characterize toxicity of oil products and effi- cacy of remediation. Externally-brooded embryos require 12 d incubation to hatch at test temperatures (27°C) and developmental endpoints are easy to detect through the transparent embryonic coat. By 3 d after oviposition embryo organ systems are beginning to develop, by 5 d the heart is visibly contracting, and by 8 d the compound eyes are fully developed. Developmental abnormalities due to exposure to solvents include abnormal eye formation and pigmentation, disfigured telson. diminished hepatopancreas and developmental delay. Abnormalities are seen as early as 4 d after oviposition and as late as 2 d post-hatch. The first layer of the outer envelope of the embryonic coat is secreted by the pleopods of the female. During development, three additional envelopes (comprised of at least 7 new layers) are generated from the em- bryo. By day 9. the outer envelope begins to erode as if in prep- aration for hatching. Concomitantly, the permeability of the coat is increased and the embryos become more sensitive to toxicants. This can be traced by the penetration of fluorescent beads of dif- ferent sizes. Partly due to this increased permeability, acute 96 h exposures to toxicants at this late stage of development (initiated on d 9) result in LC^,, values relatively close to those generated by the 12 d tests. For example, the 12 d LC50 for water soluble fraction of fuel oil is 15.2% v/v and the 96 h LC50 is 2239c. Comparatively close values have also been obtained for ethanol. DMSO and acetone. HISTOPHAGOUS CILIATE DISEASES OF CRUSTACEA. J. Frank Morado, National Oceanic & Atmospheric Administra- tion. National Marine Fisheries Service. Alaska Fisheries Science Center, 7600 Sand Point Way NE. Seattle, WA 981 15. In 1888 the first histophagous ciliate infection of a crustacean was reported in Europe. At the time of this discovery, the com- plete absence of circulating host hemocytes was noted: this obser- vation eventually became the hallmark of histophagous ciliate in- fections of crustaceans. It was also noted that the host was a recently molted crustacean, but this observation was consider of lesser importance. Histophagous ciliate infections have since been sporadically reported in both wild and captive crustaceans from the eastern North Pacific, and the western and eastern North Atlantic Oceans. Until recently all histophagous ciliate infections were re- stricted to marine crustaceans. Until 1980. the sporadic and infrequent reports of systemic ciliate disease of crustaceans and their low prevalences suggested that ciliate infections could be of importance in aquaculture, but not in the population dynamics of wild Crustacea. Since 1980. the evidence now clearly indicates that systemic ciliate disease is a major problem in the culture of the Australian crayfish, Cherax quadncarinatus . and in the captive maintenance of the American lobster. Hoinarus americanus. and Dungeness crab. Cancer mag- ister. The evidence further indicates that systemic ciliate disease may play an important role in the population dynamics of wild Dungeness crab and the American lobster. Almost from the time of their initial discovery, the taxonomic identity of the disease causing agents has been a major point of confusion. Over the last couple of years however, considerable effort has been directed toward the identification of the respective ciliated protistans. This presentation will discuss the taxonomic affinities of the respective disease causing agents, some aspects of disease pathogenesis, and review the history and present a current perspective of histophagous ciliate infections of crustaceans. EFFECTS OF DIMILIN ON THE BLUE CRAB, CALLI- NECTES SAPIDUS IN SHALLOW WATER HABITATS. Steve Rebach, Department of Natural Sciences, University of Maryland Eastern Shore, Princess Anne, MD 21853. Diflubenzuron (Dimilin) is used as an insecticide for Gypsy Moth control. Because it is a chitin synthetase inhibitor, it can be a potential threat to other arthropods including crustaceans. We used a static renewal testing paradigm to determine the LC^,, of Dimilin WP-25 to juvenile blue crabs (carapace width: 25 mm-60 mm). Both molt stage and dose frequency affected toxicity. When we exposed the crabs at random molt stages, LC5n = 3.5 mg 1 ~ ' . When crabs were exposed on the day of molt, LC51, = 300 (j,g 1 ' . If initial exposure occurred on the day of molt and the crabs were subsequently exposed to repeated doses, LC5Q = 18.5 |i.g 1~ '. Effects were age and molt stage sensitive. THE PARASITIC DINOFLAGELLATES OF MARINE CRUSTACEANS. Jeffrey D. Shields, Chesapeake Bay National Estuarine Research Reserves in Virginia, Virginia Institute of Ma- rine Science, Gloucester Point, VA. 23062. Dinoflagellates are often thought of as free-living, autotrophic protistans that live in pelagic or neritic surface waters. There are. however, many heterotrophic and parasitic forms. The latter in- clude some unusual parasites of crustaceans that are relatively unknown. There are two orders that parasitize crustaceans: the Blastodinida. with two families, and the Syndinida. with one fam- ily. Parasitic dinoflagellates infect copepods. amphipods. mysids. euphausiids. and decapods. They inhabit the eggs, stomach, soft tissues and the hemal sinuses of their hosts. Their pathogenicity varies with their invasiveness in the host. The gut-dwelling blas- todinids are relatively benign, while the chytriodinids kill their host egg. Members of the pervasive Syndinida are overtly patho- genic and insidiously ramify throughout the hemal sinuses and organs of their hosts. Host castration, feminization, lethargy, and eventually death are common results of infection. Past work ex- 490 Abstract. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore, Maryland amined the taxonomy, nuclear organization and division, and host- parasite relationships of these parasites. More recently studies have compared ultrastructural differences between parasitic and free-living forms. With the advent of major epizootics of Hema- todinium sp. on commercially important crustaceans, and out- breaks of Syndiniiim and Btastodiiuuin spp. on copepod popula- tions, attention has shifted to the economic and ecological impacts of these parasites on their host populations. Recent epizootics have caused significant financial losses to the afflicted commercial fish- eries. These epizootics appear associated with host-parasite sys- tems that occur in regions with unique hydrological features, such as fjords or poorly draining estuaries with shallow sills. These regions are ideal for the application of a "landscape ecology" approach that could lead to a better understanding of the epizooti- ology of parasitic dinotlagellates and other marine pathogens. ECOLOGICAL FUNCTION OF BIVALVES INTERTIDAL OYSTER REEFS AS CRITICAL ESTUA- RINE ENVIRONMENTS: EVALUATING HABITAT USE, DEVELOPMENT AND FUNCTION. Loren D. Coen,* Eliza- beth L. Wenner, David M. Knott, Bruce VV. Stender, Nancy H. Hadley, and M. Yvonne Bobo, Murine Resources Research Institute. South Carolina Department of Natural Resources. Charleston. SC 29412. In South Carolina, over 95% of the oysters grow intcrtidally, forming extensive "biogenic" reefs that are very different from oyster reefs studied elsewhere. Whether these intertidal habitats are functionally equivalent to other structurally complex habitats, such as seagrasses or salt marshes, is an important question. We have constructed three replicate experimental reefs (each 23 m") at each of two sites: a "pristine" oyster Hat and a "degraded" area adjacent to a marina. A novel flume net system was developed for quantifying transient species, and last spring we initiated sampling of the transient fauna and the resident reef community on each experimental reef and on an adjacent natural reef of equivalent size. We are also collecting continuous environmental data and comparing contaminant levels, oyster diseases (Dermo and MSX) and other life history parameters (e.g.. growth, condition indices, reproduction) on the experimental and natural reefs. To date we have collected over 38 species of transient fishes and decapod crustaceans (dominant genera, Penaeus, Palaemonetes, Anchoa, Leiostomus. Cobiosoma). with peak densities exceeding 2,400 individuals/reef. Samples of resident fauna have been sorted and identified only from the first season; preliminary analyses show no differences between sites or between experimental and natural reefs, with regard to overall abundance or species richness. Indi- vidual species, however, showed marked differences in abundance ;iattems. The oyster parasite Boonea impressa. for example, nu- merically dominated the natural reef resident community, but was not found in the initial sampling of the experimental reefs. The complete temporal series of samples will be used to explore changes in the status of reef habitat and function throughout its succession. OYSTER PRODUCTION— A LARGE SCALE PERSPEC- TIVE. Jerry McCormick-Ray,* Dept. Environmental Sciences. Clark Hall, University of Virginia, Charlottesville, VA 22903. Oysters have evolved with the Chesapeake Bay ecosystem. Historical evidence supports their widespread distribution and abundance, forcing consideration of the intergration of the oysters' biology with the physical system. Ecological science suggests the critical role of Crassostrea virginica in shoreline changes; in marsh development, erosion, and sediment deposition; in food chains and nutrient transfers; in water quality, hydrodynamics, and benthic function. These ecological roles are important to a diverse array of species, including other species of commercial impor- tance, and they are important to ecosystem change. A large scale perspective of the oyster in the Chesapeake Bay highlights the need to focus oyster productivity and sustainment on watershed management and ecosystem processes. PHYSICAL AND BIOLOGICAL SCALING OF BENTHIC- PELAGIC COUPLING IN COASTAL ECOSYSTEMS: THE ROLE OF BIVALVE SUSPENSION FEEDERS. Elka T. Por- ter,* Roger I. E. Newell, and Lawrence P. Sanford, Horn Point Environmental Laboratory. University of Maryland. Box 775, Cambridge. MD 21613. It is increasingly recognized that benthic suspension-feeding bivalves are an important component of estuarine ecosystems be- cause they increase the transfer of seston and particle-bound tox- icants from the water column to the benthos and because of their role in nutrient regeneration. Water flow is crucially important for supplying seston to the bivalves and controlling benthic-pelagic nutrient fluxes at the sediment-water interface. Previous studies on seston removal by bivalves and nutrient regeneration have been performed either in flumes in the labora- tory or field flumes, and in laboratory mesocosms such as the Marine Ecosystems Research Laboratory at the University of Rhode Island. A disadvantage of tlumes is that water column processes are not adequately scaled although benthic boundary- layer processes are well represented. Mesocosm water column processes may be represented well but processes at the sediment- water interface are not accurately scaled. Minimal mixing at the sediment-water interface in mesocosm tanks is problematic for accurately studying benthic processes because low water flow di- rectly affects seston uptake by bivalves and may lead to refiltra- tion. enhanced sedimentation, reduced resuspension. and a change in nutrient regeneration processes. Therefore, results from such studies in flumes or mesocosms cannot be scaled up to ecosystem level with confidence. We utilized coupled mesocosm and flume systems to study nutrient and particle fluxes at the benthic boundary layer in order National Shcllfisheries Association. Baltimore. Mai-vland Ahsiract. 1996 Annual Meeting. Apnl 14-18. 1996 491 to obtain benefits inherent in both experimental systems. We stud- ied the effect of the interaction of benthic bivalve suspension feed- ers and water flow on seston quantity and quality and on nutrient regeneration in systems of two different sizes and with different water flows at the sediment-water interface. Preliminary results of replicate experiments indicate that oysters (C. virginica) decrease Chi a and seston concentrations in systems of both sizes, but water flow did not appear to have a substantial effect. However, water flow at the sediment-water interface did significantly affect nutri- ent regeneration in systems with bivalves and in systems of both sizes. These results emphasize the importance of considering wa- ter flow in ecosystem studies with benthic bivalve suspension feeders. TROPHIC COMPETITION BETWEEN THE PACIFIC OYSTER CRASSOSTREA GIGAS AND THE POLYCHAETE LANICE CONCHILEGA IN THE BAY OF VEYS (FRANCE). Michel Ropert, IFREMER. Shellfish Aquaculture Laboratory. B.P. 32. Port en Bessin; P. T. Goulletquer,* IFREMER. GAP/ URAPC Laboratory. B.P. 133. 17390 La Tremblade. France; J. P. Joly, IFREMER Port en Bessin. The Bays of Veys leasing grounds (i.e.. 406 acres), represent- ing a yearly oyster production of 10.000 metric tons, are fully exploited. Strongly spatially correlated to these leases and inter- fering with the oyster industry, a population of the polychaete Lattice cottchilega has drastically increased since 1986 to reach 8.000 individuals.m"" in several areas. The trophic competition between the Pacific oyster Crassostrea gigos and the polychaete L. conchilega was studied by assessing the polychaete suspension feeding activity at the laboratory. Particle size distributions were compared at the input and output of an experimental design to estimate the polychaete retention efficiency. Although particles ranging from 4 p.m to 12 (im were kept by L. conchilega. no upper threshold and maximum retention rate were reached. In contrast. C. gigas showed a retention efficiency at 2 ixm and reached a 6 to 8 (xm upper threshold. Based on our results, a trophic competition is likely to occur between C. gigas and L. conchilega, therefore affecting the oyster industry. Standardized filtration rates reached 0.27l.h"'.dmw"' and 2.41 l.h^'.dmw"' ford g) L. conchil- ega and C. gigas respectively. Polychaete assimilation rates (0.37) were significantly lower than those of C. gigas (0.49). Respiration rates were estimated to 0.21 and 0.62 ml.O^.h '.dmw ' forL. conchilega and C. gigas respectively. Therefore, polychaete scope for growth (SFG) (2.17 J.h^'.dmw"') was significantly lower compared to the 63.7 J.h'Vdmw ' C. gigas's SFG. Based on these results and both species field stock assessments, L. conchi- lega was responsible for 21% carrying capacity decrease. How- ever, L. cottchilega' s SFG represented only 5% of C. gigas pop- ulation's SFG. Several hypothesis regarding both populations in- teractions and further management are discussed with regards to physical and biological assumptions. ECONOMICS OF THE AQUACULTURE INDUSTRY CHARACTERISTICS OF ON-BOTTOM OYSTER RACK STRUCTURES IN THE CHESAPEAKE BAY. Eric J. Powell, 805 Buckingham Drive, Stevensville, MD 21666. The drastic decline in Maryland's oyster harvests over the past several decades has resulted in a change of direction in the man- agement of oyster stocks including the implementation of experi- mental off-bottom oyster culture leases. This project incorporates the results of previous disease research and off-bottom rack studies to develop the characteristics of an on-bottom rack culture system. This system was designed to be worked using traditional com- mercial fishing vessels and techniques. Results from this study found that there was no significant difference in growth between oysters raised in a rack structure and those grown in floating trays. This study was performed at a low salinity (6 weeks. Ventilation and survivor- ship were then measured as described above. Our preliminary data indicates that coastal animals held at estuarine salinity levels show improved osmoregulatory capabilities. These data suggest that dif- ferences in the osmoregulatory capabilities of coastal and estuarine lobsters are probably due to acclimation, which in part, induces expression of increased Na*/K* ATPase activity. TEMPERATURE CONTROL OF RECRUITMENT IN THE AMERICAN LOBSTER. S. L. Waddy and D. E. Aiken, In- vertebrate Fisheries. Maritime Region. Dept. Fisheries and Oceans. St. Andrews. NB Canada EGG 2X0. The American lobster experiences a wide range of thermal conditions within its natural range and experimental studies have led to many advances in the understanding of how temperature controls growth and reproductive functions. Temperature control mechanisms are complex as the various environmental factors reg- ulating lobster biology often act synergistically. resulting in re- sponses that are difficult to predict if the various factors are studied in isolation. Several intriguing mechanisms involving temperature and other environmental regulators have been identified. It is known that temperature, season and daylength interact to control larval development, juvenile and adult growth, and maturation and reproduction. Dramatic physiological changes occur at the autum- nal equinox and the winter solstice (and probably the spring sol- stice and summer equinox as well) and many biological processes appear similarly affected by real calendar time. For example, the temperature required to induce premolt and ovarian maturation is considerably lower in the spring than in the autumn and a differ- ence of only a few days at critical times of the year can produce diametrically different responses to temperature. The rate of ovar- ian maturation varies from a few weeks to a few months depending on the time of year, and larval development at a given temperature can vary by more than SC/t between late spring and late summer. Many growth and reproductive processes require that the temper- ature reach a certain threshold and the threshold changes with both daylength and time of year. Accumulating evidence indicates that the thermal requirements for development, growth and reproduc- tion are not cumulative (as in degree-days) but rather are a com- bination of threshold and cumulative phenomena. MOLLUSCAN DISEASE I STATUS OF OYSTER DISEASES IN MARYLAND'S OYS- TER RECOVERY AREAS. Gustavo W. Calvo* and Stephen J. Jordan, Maryland Department of Natural Resources. Cooper- ative Oxford Laboratory, 904 S. Moms St., Oxford, MD 21654. Oyster recovery areas (ORA) currently regulate the placement of seed and harvest of oysters by zoning Maryland's rivers. We report on the status and variability of diseases in Buoy Rock (CHBR) and Old Field (CHOP) in Chester River-zone "C"; Cabin Creek (CRCA). Oyster Shell Point (CROS). and Irish Creek (BCIC) in Choptank River-zones "A". "B". and "C". respec- tively; and Wilson Shoals (NRWS) in Nanitcoke River-zone "C". Sampling was conducted once seasonally in spring, summer, and fall of 1995 by collecting two replicate samples of 30 oysters 496 Abstract. 1996 Annual Meeting, April 14-18. 1996 National Shellfisheries Association. Baltimore, Maryland in each oyster-bar. Temperature and salinity were recorded at the time of sampling. Diagnosis for P . marinus was by Ray's fluid thioglycoUate medium assay of hemolymph and diagnosis for H . nelsoni was by hemolymph analysis and paraffin histology. As expected, prevalence and intensity of P. inunnns infections increased as the seasons progressed. In the fall, prevalence was 73% in CHOP (12 ppt). and 307f in CHBR (12 ppt); 30% in CRCA (10 ppt), 97% in CROS (12 ppt). and 98% m BCIC (15 ppt); and 76% in NRWS (14 ppt). The above results represent a 19%-50% increase in salinity and a 30%-85% increase in fall prevalence relative to values recorded in 1994. Variation in prev- alence between the two samples collected from each oyster-bar was <30% (N = 20). Haplosporidium nelsoni was present in Nanticoke River oysters. We will discuss these results in relation to changing environmental conditions and management implica- tions. PERKINSUS MARINUS TRANSMISSION DYNAMICS IN CHESAPEAKE BAY. Lisa M. Calvo* and Eugene M. Burre- son, Virginia Institute of Marine Science. College of William and Mary, Gloucester Point, VA 23062: Christopher F. Dungan, Cooperative Oxford Laboratory, Maryland DNR, Oxford, MD 21654; Bob S. Roberson, University of Maryland. College Park, MD 20742. This study is the first to systematically examine the seasonality oi Perkinsus marinus transmission in eastern oysters. Crassostrea virginica. in relation to water column abundance of P. marinus cells, host oyster mortality, and temperature. The study was con- ducted for a one year period in the lower York River, VA. Flow cytometric immunodetection methods were employed to determine P. marinus cell abundance in water samples collected 3 times a week. Uninfected sentinel oysters were deployed biweekly to de- termine infection transmission rates and local host oyster mortality rates were estimated biweekly by monitoring a captive population of native oysters. Environmental abundance of P. mariiuis cells, infection acqui- sition by sentinel oysters, and mortality of P. marinus infected oysters varied seasonally. Distinct peaks of all three parameters occurred during the month of August, following maximal summer temperatures. Water column parasite cell abundances, infection pressure, and oyster mortalities decreased from summer maxi- mums as temperatures decreased in September and October and remained at low levels from October through the termination of the study in March. Strong and significant positive correlations were found between water column parasite cell abundance and temper- ature, water column parasite cell abundance and oyster mortality, oyster mortality and temperature, and oyster mortality and P. marinus prevalence in sentinel oysters. Perkinsus marinus preva- lence in sentinel oysters did not significantly correlate with water column cell abundance. These results support the currently ac- cepted hypothesis that infective stages of P. marinus originate from dying oysters and are most abundant in August. The occur- rence of elevated cell abundances in early summer, immediately before epizootic oyster mortalities, suggests that pathogen cells may also originate from other sources. USE OF SPECIFIC-PATHOGEN-FREE (SPF) OYSTERS TO MEASURE GROWTH, MORTALITY, AND ONSET OF MSX AND DERMO DISEASE IN SOUTH CAROLINA. Nancy H. Hadley,* M. Yvonne Bobo, Donnia Richardson, and Loren D. Coen, Marine Resources Research Institute, Charles- ton. SC 29412; David Bushek, University of South Carolina, Belle Baruch Institute, Georgetown. SC 29440. The SCDNR's Marine Resources Research Institute has re- cently initiated a long-term study of intertidal oyster reef ecology, utilizing two field sites, a "degraded" marina and a "pristine" tidal cieck. Native oysters at these sites have been monitored monthly since September 1994 and both Dermo and MSX have been detected at prevalences as high as 100% (Dermo) and 42% (MSX). As part of this study we planted Specific-Pathogen-Free (SPF) oysters at these sites to monitor growth, mortality, and onset of MSX and Dermo diseases. These SPF juveniles were produced from native intertidal stocks in our hatchery and reared under strict quarantine. All seawater was filtered to 0.45 |j.m and UV irradi- ated to remove or destroy infective stages present in our local water supply. Juveniles spawned in March 1995 were subsampled (n = 50) at 4 months of age to determine disease status. Assays for MSX (histological method) and Dermo (RFTM body burden method and polyclonal antisera reactions) were negative. Individ- uals averaging 12.3 mm shell height (SH) were planted at the two field sites in July 1995 in plastic mesh bags (200 oysters/bag, 3 bags per site). Subsamples (10 oysters/bag, 30/site) were assayed after I. 2. 4, 8 and 12 weeks of exposure. Dermo was first de- tected at the marina site at 8 weeks and at the tidal creek site at 12 weeks. MSX was not detected in the deployed oysters during the first 3 months of the study, which will continue through spring 1996. After 4 months of field exposure, oysters averaged 47.8 mm SH at the tidal creek site and 43.4 mm at the marina site (p = 0.07) and mortality was low (6.5-11.5%). DISTRIBUTION AND POPULATION DYNAMICS OF A HYDROZOAN INQUILINE SYMBIONT OF THE EAST- ERN OYSTER. Dale S. Mulholland* and Frank E. Friedl, Department of Biology. University of South Florida, Tampa, PL 33620. My investigations of a cnidarian symbiont lightly attached to the gills and mantle of the eastern oyster. Crassostrea virginica, raise questions of their impact on oyster health and human con- sumers. A survey of intertidal oysters in the Gulf of Mexico from Grand Terre. LA. to Tampa Bay. FL. and in the Atlantic from Charleston, SC, to Palm Beach County, FL. found no cnidarians inoystersnorthof Tampa bay or north of St. Augustine, FL. Many locations in Tampa Bay and along the central Atlantic coast of Florida yielded oysters with these symbionts. National Shcllt'ishcries Association. Baltimore. Maryland Abstract. 1996 Annual Meeting. April 14-18, 1996 497 Previously, immature specimens from Ft. Pierce. FL, were placed in the genus Entinui by Kubota and Larson (Proc. of Jap- anese See. of Syst. Zool. no. 41. 1990). More details of the morphology of polyps and medusae are now available. Monthly collections from intertidal oyster populations in Tampa Bay were made for a year, and the results suggest some explanations for the apparent geographic distribution. Cnidarian numbers dropped drastically after oysters were exposed to freezing temperatures during low tide. The cnidarian population was also severely depressed during a summer of unusually heavy rainfall and reduced water osmolality. Population dynamics of this hydrozoan suggest they are capa- ble of rapid asexual reproduction when conditions are favorable; hundreds, even thousands, can be found in any one oyster. These large numbers raise concerns about the use of affected oysters as food. Sexually reproducing medusae soon appear if conditions remain favorable. However, the number of oysters with cnidarians does not increase rapidly, suggesting the cnidarians have limited or vulnerable means of distribution. PERhL\SLS MARIM'S TISSUE DISTRIBUTION AND SEA- SONAL VARIATION IN OYSTERS (CR.ASSOSTREA VIR- GINICA) FROM FLORIDA. VIRGINIA AND NEW YORK. Leah M. Oliver* and W. S. Fisher, U.S. EPA National Health and Environmental .Effects Laboratory, Gulf Ecology Division. 1 Sabine Island Drive. Gulf Breeze, FL 32561-5299; E. M. Bur- reson and L. M. Ragone-Calvo, Virginia Institute of Marine Sci- ence, Gloucester Point, VA 23062; S. E. Ford and J. Gandy, Haskin Shellfish Research Laboratory, Institute of Coastal and Marine Sciences, Box B-8, Rd. #1 . Port Norris, N.J.. 08349-9736. Perkinsus marintis infection intensity was measured in Amer- ican oysters (Crassostrea virginicu) that were collected in October 1993. December 1993, March 1994, May 1994 and July 1994 from Apalachicola Bay (Florida). Chesapeake Bay (Virginia), and Oyster Bay (New York). Gill, mantle, digestive gland, adductor muscle, hemolymph and other remaining tissue (including gonadal material) were dissected from 20 oysters from each site at each collection time, and were separately diagnosed for Perkinsus mari- nus infections by incubation in Ray's Fluid Thioglycollate Media and subsequent microscopic quantification of enlarged prezoospo- rangia. At all sampling times and sites, average P. marinus infec- tion intensity (# parasites gm" ' wet weight tissue or ml ' he- molymph) was consistently lowest in hemolymph samples, and the greatest density of the parasite was generally found in the digestive gland. P. marinus prevalence was 100% at all sites for each of the 5 collection times. Seasonal infection intensity patterns and mean total oyster body burdens differed among the sites. The highest average body burdens were measured in Virginia oysters in Sep- tember 1993 but were lower by December probably due to mor- tality of heavily infected oysters and dimmution of parasite activity typically associated with colder temperatures. During the same time period, infection mtensities remained at about the same level or slightly higher for Florida and New York oysters, respectively. The lowest infection intensities were measured in March for Flor- ida oysters, and in May for oysters from Virginia and New York. The retention of relatively high P. marinus infection levels in New York oysters compared to Virginia oysters may have been due to constant high salinity in Long Island Sound plus a very rapid decline in temperature in the fall which may have prevented epi- zootic mortalities such as those often associated with P . marinus. PHYLOGENETIC POSITION OF THE GENUS PERKINSUS BASED UPON ACTIN GENE SEQUENCES. Kimberly S. Reece,* Mark E. Siddall, and John E. Graves, Virginia Institute of Marine Science, College of William & Mary, Gloucester Point. VA 23062. Perkinsus species are presently classified within the phylum Apicomplexa. This placement, however, is controversial. Based upon both morphological observations and nucleotide sequence data of the small subunit rRNA gene some have suggested that Perkinsus species may be more closely related to dinoflagellates than to apicomplexans. To reevaluate the phylogenetic position of Perkinsus. we obtained nucleotide sequence data for actin genes from Perkinsus marinus, a haplosporidian and three species of dinoflagellates. DNA was isolated from in vitro cell cultures of P. marinus. cultures of three dinoflagellates. Gyrodinium uncate- num. Gymnodinium sanguineum and Prorocentrum minimum and from the spores of Haplosporidian louisiana. Actin gene frag- ments (622 bp) were amplified from each of the genomic DNA isolations using "universal" actin gene primers in the polymerase chain reaction (PCR). Genomic Southern blot analysis indicated that there are two closely related actin genes in P. marinus. The two actin fragments amplified from P. marinus genomic DNA have been sequenced and demonstrate 97.4% nucleotide sequence identity. All of the 16 observed nucleotide differences occur at the third position of codons and the encoded protein fragments of these two genes are identical. Amplified fragments from the di- notlagcllates and the haplosporidian are presently undergoing DNA sequence analysis. For phylogenetic analyses, several pro- tozoan actin gene sequences, in particular those from apicomplex- ans and ciliates, were downloaded from the Genbank (NCBI) se- quence database. Parsmiony analysis with these actin gene se- quences will be done to constuct phylogenetic trees. PERKINSUS MARINUS, FLOW CYTOMETRIC IMMUNO- ASSAY AND INTERANNUAL ABUNDANCE IN CHESA- PEAKE BAY ESTUARIES. Bob S. Roberson* and Tong LI, University of Maryland. Department of Microbiology, College Park, MD 20742; Christopher F. Dungan, Cooperative Oxford Laboratory, Oxford, MD 21654; Eugene M. Burreson, Virginia Institute of Marine Sciences, Gloucester Point, VA 23062. The development of flow cytometric immunodetection meth- ods for enumerating Perkinsus marinus in the diverse mix of par- ticulates found in water samples from estuarine waters of the Ches- 498 Abstract. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore. Maryland apeake Bay, has offered a means for monitoring the seasonal en- vironmental abundance of this oyster pathogen. The method employs antibody produced in rabbits against cultured forms of the organism, to permit specific fluorochrome labeling of Perkinsus sp. cells. The immunolabelled cells in water sample particulates were detected by flow cytometry. By adjusting trial count ranges of fluorescence, size, roughness, nucleic acid content or autoflu- orescence. followed by sorting from a sample pool made from the previous year's set of samples, it was possible to select ranges for each parameter that would exclude contaminating non-Peikinsus particulates arising or anticipated during the year. Parameters were adjusted until the sort was judged by microscopic examination to be free of contaminating particulates. This technique has been used to monitor environmental abundance of the pathogen in the Tred Avon River at Oxford, Maryland for three years, and from a second, higher salinity site, the mouth of the York River, during the last year. Although no period of the year yields samples free of Perkinsus organisms, abundance peaks occur during the summer months from June through August. HEMATOLOGICAL PATHOLOGY OF WASTING SYN- DROME IN BLACK ABALONE. Jeffrey D. Shields* and Frank O. Perkins. Virginia institute of Marine Science. The College of William and Mary. Gloucester Point, VA 23062; Car- olyn S. Friedman, Calif. Fish & Game, Fish Health Laboratory. Bodega Mannc Laboratory. P.O. Box 247. Bodega Bay. CA 94923. Withering syndrome is a debilitating and fatal disease of the black abalone. Haliotis cracherodii . The etiological agent for the disease is presently unknown. Statistical analyses of hemocyte counts, and various parameters from the stained blood smears of over 600 abalone indicate a pattern of disease in the blood of the host. We tentatively identify three hemocyte types as representa- tives of a single class of lymphocyte. Small lymphocytes are most likely stem cells, that develop into hyalmocyte-type cells, that grow further into large fibrocyte-like cells. Small lymphocytes were not abundant in the lymph of healthy abalone (35.4% prev- alence), but were common in wasted animals [11.9% prevalence). Hyalinocytes and fibrocytes declined sharply in abundance in wasted animals. In wasted animals, fibrocytes dehisce almost im- mediately upon contact with a microslide. Fibrocytes from healthy animals typically dehisce after 10+ minutes. In addition, healthy abalone had a mean of 1.73% dead cells in their blood compared to a mean of 4.69% in slightly shrunken abalone. and 10.32% in the blood of wasted abalone. Cellular inclusions and an unusual cell type were also significantly more prevalent in wasted than in healthy hosts. Cellular inclusions were found in 40% of wasted abalone compared with 15% in nonsymptomatic hosts. (Note that the prevalence in the healthy animals may represent subclinical infections.) The unusual cell type was found in 28% of wasted abalone compared with 15% in nonsymptomatic hosts. There was also a strong positive relationship between the presence of the unusual cell type and the presence of cellular inclusions. Our results support the view that the infectious agent may be blood borne, and that it causes lysis of hemocytes. COMPARISON OF HAPLOSPORIDWM NELSONI DIAG- NOSTIC TECHNIQUES: POLYMERASE CHAIN REAC- TION OUTPERFORMS HISTOLOGY. Nancy A. Stokes,* Juanita G. Walker, and Eugene M. Burreson, Virginia Institute of Manne Science. College of William and Mary. Gloucester Point. VA 23062. Histological examination of preserved, sectioned, and stained oyster tissue has been the method of choice for detection of the protistan pathogen Haplosporidium nelsoni for the past 30 years. We recently developed primers for polymerase chain reaction (PCR) which are sensitive and specific for the small subunit ribo- somal DNA of W. nelsoni. Detection of H. nelsoni-infected oys- ters by the techniques of histological examination and PCR were compared in monthly samples of oysters held in trays in the York River, VA over an eight month period beginning May 1995. PCR was conducted on DNA extracted from both hemolymph and gill and mantle tissue and histological examination was conducted on transverse sections through the visceral mass including gill and mantle. Haplosporidium nelsoni was not detected by histology or PCR of tissue DNA until July, whereas PCR of hemolymph DNA de- tected low prevalence levels in May and June. In the July. August, and September samples. PCR of tissue DNA detected almost twice as many infections as histological examination, while PCR of hemolymph DNA detected as many or more infections as histo- logical examination, but fewer than PCR of tissue DNA. Of the 122 oysters already screened by both techniques. 42 H. nelsoni infections were detected by PCR but not by histological examina- tion, whereas 2 local infections were detected by histological ex- amination but not by PCR. MOLLUSCAN DISEASE II THE EFFECT OF PENTACHLOROPHENOL ON NADPH PRODUCTION IN OYSTER HEMOCYTES: IMMUNO- MODULATORY CONSEQUENCES. Cal Baier-Anderson* and Robert S. Anderson, Chesapeake Biological Laboratory. University of Maryland. P.O. Box 38. Solomons. MD 20688. The generation of reactive oxygen species (ROS) by Crassos- trea virginica hemocytes is thought to be an important line of defense against pathogens. ROS are produced by NADPH oxidase during the respiratory burst, when an electron from reduced 3-nic- otinamide adenine dinuclcotidc phosphate (NADPH) is transferred to oxygen, forming the superoxide anion. Since increased produc- tion of NADPH is required for ROS generation, one possible National Shellfisheries Association, Baltimore. Maryland Abstract. 1996 Annual Meeting. April 14-18. 1996 499 mechanism of immunotoxicity is to inhibit NADPH production. The effects of a putative environmental immunotoxicant on NADPH production and the concomitant effects on ROS modula- tion are reported here. Oyster hemocytes were exposed in vitro to sublethal concen- trations of pentachlorophenol (PCP). a biocide and uncoupler of oxidative phosphorylation that inhibits ROS generation, and as- sayed for increased NADPH production following stimulation with phorbol myristate acetate. NADPH production was estimated using a colorimetric method that measures intermediate metabo- lism in terms of the reduction of 3-(4.5-dimethylthiazol-2-yl)-5- (3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium, in- ner salt (MTS) via phenazine methosulfate (PMS). In this assay. electrons are transferred from NADPH to MTS in a series of redox couples where PMS is the intermediary. Preliminary results indi- cate that PCP partially inhibits the production of NADPH in a dose-dependent manner, which could explam decreased ROS pro- duction. The implications of decreased NADPH production in terms of toxicity and immunocompetence will be discussed. ISOLATION OF A cDNA CLONE FROM MERC EN ARIA MERCENARIA THAT CODES FOR A PROTEIN RELATED TO THE CYTOCHROME P4S0 III SUBFAMILY OF EN- ZYMES. David Brown,* Duke University Marine Lab, Beau- fort. NC 28516; George Clark, National Institute of Environmen- tal Health Safety. Research Triangle Park. NC 27709; Rebecca Van Beneden, University of Maine. Orono, ME 04469. We are using molecular techniques to identify possible factors in the high prevalence of gonadal tumors observed in Mya are- naria from Washington County Maine and Mercenaria merce- naria from the Indian River in Florida. The etiology of the high prevalence of tumors is not known, but genetic factors and expo- sure to agricultural chemicals seem to be involved. It appears that gender might also play a role. In both of the bivalve populations the prevalence of tumors was highest in females and female clams tended to have more aggressive, invasive tumors. As part of our investigation into the molecular mechanisms for the etiology of these tumors, we have isolated several cDNA clones from a M. mercenaria cDNA library. One clone was se- quenced and the predicted amino acid sequence aligns with known sequences for vertebrate cytochrome P450 enzymes. The highest amino acid homology was to enzymes in the cytochrome P450 III subfamily. This is one of the older cytochrome P450 subfamilies. and in vertebrates this family of enzymes is involved in the me- tabolism of steroids. Steroids have been implicated in the devel- opment of tumors in vertebrates and information from the epi- zootic studies suggested sex hormones or at least gender could be involved in the development of gonadal tumors in M . arenaria and M. mercenaria. Therefore, the isolation of this cytochrome P450 clone has opened up several new areas of investigation into the development of these gonadal tumors as well as the role of cy- tochrome P450 enzymes. INFECTIVITY AND PATHOGENICITY OF PERKINSUS MARINUS. 3. FECAL ELIMINATION. David Bushek*, Baruch Marine Laboratory/USC, PC Box 1630. Georgetown. SC 29442; Susan E. Ford and Marnita M. Chintala, Haskin Shell- fish Research Laboratory. Rutgers University, IMCS. Box B-8, Port Norris. NJ 08349. The rate at which hosts shed live parasites has numerous im- plications for understanding a host-parasite relationship. It may reflect I ) the initial host response to an experimentally delivered dose; 2) differential reaction to qualitative parasite differences; 3) infection development rate; 4) actual mfection intensity; and 5) transmission potential. We examined the interaction between Per- kinsus marimts and Crassostrea virginica by measuring the rate at which parasites appeared in the feces and pseudofeces of experi- mentally infected oysters at various times post challenge. The daily discard rate decreased during the week following challenge. Initially, more cells were found in the pseudofeces than in the feces, but by day 7 relatively few parasites were found in the pseudofeces. After day 7. the rate of discard in feces increased exponentially, indicating that parasites in the original dose were probably shed in decreasing number for about I week; thereafter, increasing numbers mirrored the intensification of infections over time. This explanation is supported by the finding that the number of parasites shed was roughly correlated with time to death. The continuous release of viable parasites from live oysters represents an important potential transmission mechanism in nature and may provide a nondestructive method of estimating infection intensity. When oysters were challenged with equal numbers of cultured or wild parasites from naturally infected oysters they eliminated 10 times more cultured than wild cells in the feces during the first two weeks post-inoculation. The pseudofeces ratio was 50; 1 (cultured: wild) suggesting that some qualitative difference between the par- asite types favored retention of wild cells or elimination of cul- tured cells by the gills and palps. There was no difference in the number of lag. log, and stationary phase cultured parasites elim- inated during the first day post-challenge, indicating that size or stage are unlikely to account for the differences between wild or cultured cells. INFECTIVITY AND PATHOGENICITY OF PERKINSUS MARINUS. 1. PARASITE CHARACTERISTICS. Marnita M. Chintala,* Haskin Shellfish Laboratory. Rutgers University. IMCS. Box B-8. Port Norris, NJ 08349; David Bushek, Baruch Marine Laboratory/USC, PO Box 1630, Georgetown, SC 29442; Susan E. Ford, Haskin Shellfish Research Laboratory. Rutgers University. IMCS. Box B-8, Port Norris, NJ 08349. To help define the physiological characteristics of in vitro cul- tured Perkinsu.'i marinus, we assessed the infectivity and pathoge- nicity of the cultured parasites 1) in comparison to wild parasites isolated from naturally infected oysters, 2) after increasing culture passage, and 3) in different stages of cultures [lag, log, and sta- tionary]. Oysters were inoculated via the shell cavity with a weight 500 Abstract. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association, Baltimore, Maryland standardized dose of parasites and held for 12 weeks at 28°C and 25 ppt. Mortality was assessed daily and total parasite burdens were determined in all dead and surviving oysters. Most oysters inoculated with wild P. marinits died with heavy infections before the end of the expenmental period while mortality of oysters given an equal dose of cultured parasites was very low (10-13% of the wild challenge). Many oysters surviving the challenge with cul- tured parasites had infection intensities approaching those in the dead oysters, suggesting that they would have died in a longer experiment. There were no consistent differences in mortality among oysters challenged with parasites in the first, fifth, ninth, and fiftieth + culture passages (5-15% mortality). These results indicate that cultured cells are considerably less virulent than wild cells and that they lose most of their virulence during the initial culturing process. As would be expected from this, repassage of cultured P. marinus through oysters failed to restore virulence. Mortality attributable to cultured P. marinus occurred only in oys- ters inoculated with log phase parasites. Parasite burdens were highest in those challenged with log phase cells, an order of mag- nitude lower in those inoculated with lag phase parasites, and lowest in oysters dosed with stationary phase cells. IDENTIFICATION OF PERKINSUS MARINUS PORTALS OF ENTRY BY HISTOCHEMICAL IMMUNOASSAYS OF CHALLENGED OYSTERS. Christopher F. Dungan* and Ro- salee M. Hamilton, Cooperative Oxford Laboratory. Oxford. MD 21654; Eugene M. Burreson and Lisa M. Ragone-Calvo. Vir- ginia Institute of Marine Science, Gloucester Point, VA 23062. Since focal P. marinus lesions are commonly found within digestive tract epithelia of infected oysters, such epithelia have been proposed as primary portals through which infective patho- gen cells invade oyster hosts. Untested mechanistic hypotheses propose that waterbome P . marinus cells are ingested by feeding oysters, invade connective tissues via gut epithelia, and subse- quently disperse systemically to colonize all oyster tissues. In spite of available methods for experimental infection of oysters by ex- posure to a variety of P. marinus cell types, neither the proximate infective cell type(s), their invasive mechanisms, nor their portals of entry have been documented to date. We challenged uninfected oysters in the laboratory and in es- tuarine waters endemic for dermo disease. Challenged oysters were fixed, systematically subsampled histologically, and sections fluorescence-immunostained for microscopic localization of P. marinus cells and lesions. Pathogen cells were routinely observed associated with external epithelia, and within gut lumina of both experimentally- and naturally-challenged oysters. At 21 days post- exposure, experimentally-challenged oysters consistently showed lesions containing proliferating pathogen cells in epithelia and connective tissues of gill, mantle, and labial palp, and in visceral connective tissues, but not in digestive tract epithelia. These data do not support prevailing views of P. marinus infection dynamics. but suggest possible alternative mechanisms of gut epithelium col- onization which may strategically enhance parasite fitness. PROTEASE BLOCKERS INHIBIT PERKINSUS MARINUS IN VITRO AND IN VIVO. Mohamed Faisal.* and Jerome F. La Peyre, Virginia Institute of Marine Science, College of William and Mary, Gloucester Point, Virginia 23062; Craig D. Wright, Novavax Inc., Rockville, MD 20852. Recently, we demonstrated that Perkinsus marinus produces serine proteases that are important for its pathogenicity. Protease blockers are used in the control of protease-producing protozoa. In this study, we examined the effects of Bacillus lichenformis pro- tease blockers on P . marinus. The propagation of P . marinus was reduced by coincubation with the protease blockers in vitro. The degree of growth suppres- sion, however, differed between isolates. At a concentration of 10 mg/ml, the propagation of LMTX-1 isolate was totally suppressed and viability was reduced by 72%. In contrast, this same concen- tration caused a remarkable suppression in the growth rate of Per- kinsusA isolate but no mortalities. We then investigated the effects of the bacterial protease blocker on the initial phase of F. marinus infection. Oysters were injected with an overdose off. marinus ctWs (lO^/oyster) and then administered liposomes containing the protease blockers daily for 6 weeks. A significant reduction in the parasite body burden oc- curred in oysters that received the treatment (33,491 parasite/g) as compared to control oysters (328,863 parasite/g). These results suggest that protease blockers may be promising in the control of P . marinus. Further investigations have revealed that oyster plasma con- tains protease inhibitors that suppress P. marinus and other pro- teases. The involvement of plasma protease inhibitors in the de- fense against P . marinus remains to be elucidated. INFECTIVITY AND PATHOGENICITY OF PERKINSUS MARINUS. 2. DOSING METHODS AND HOST RE- SPONSE. Susan E. Ford* and Marnita M. Chintala, Haskin Shellfish Research Laboratory, Rutgers University, IMCS, Box B-8, Port Noms, NJ 08349; David Bushek, Baruch Marine Lab- oratory/USC, PC Box 1630, Georgetown, SC 29442. Host-parasite interaction studies are facilitated by the ability to infect hosts with known parasite doses under controlled condi- tions. In vitro propagation of Perkinsus marinus. a protozoan par- asite of the eastern oyster, supplies unlimited quantities of para- sites for which many characteristics are defined. To provide guide- lines for infection experiments using P. marinus and to help evaluate their results, we investigated the effect of 1) 4 different inoculation methods, 2) single vs. multiple dose feeding, and 3) the physiological condition of hosts. Oysters were given a weight standardized dose of parasites and held for 12 weeks at 28°C and 25 ppt. Mortality was assessed daily and total parasite burdens were determined in all dead and surviving oysters. Both mortality National Shellfisheries Association. Baltimore, Mainland Abstract. 1996 Annual Meeting. April 14-18, 1996 501 and parasite burdens in surviving oysters followed a pattern that roughly corresponded to the number of barriers to infection breached by the delivery method; intramuscular injection > intra- valvular injection > intubation > feeding. This pattern was found for both cultured and wild parasites, although mfections from wild parasites developed much more quickly regardless of dosing method. Oysters fed cultured P . marinus in one large dose had somewhat heavier total mortality (U'/c) than did oysters fed an identical dose that was split into 24 aliquots over 8 weeks (3%), suggesting that continuously delivered small doses (the natural situation) are easier to deal with than a single large dose (experi- mental challenge). Within two weeks of intravalvular injection. oysters with ripe gonads suffered mortalities that were much higher than in unripe oysters. The deaths were attributed to a combination of the injection, spawning, and water that became fouled with dense gamete concentrations. We found no evidence that holding oysters for a prolonged period (6 weeks) before chal- lenge altered P. mariniis-OMxA mortality. OYSTERS, OXYGEN METABOLISM. AND HEMOCYTES. Frank E. Friedl,* Department of Biology, University of South Florida, Tampa. FL 33620-5150. Bivalves in general are considered to be succinate-producing facultative anaerobes able to accommodate hypoxia. They have metabolisms with both aerobic and anaerobic components. Using microelectrode techniques similar to those reported for the fresh- water clam EUiptio (Vitale and Friedl 1994. J. Shellfish Res. I3(l);301), we have found surfaces and tissues of monovalve Crassostrea virginica preparations in normoxic environments deeply hypoxic with ambient surface gradients indicating oxygen uptake. The animals appear to be dependent on whole body dif- fusion for gas exchange, and have a "tissue-animal" character. We have shown that Crassostrea hemocytes take up oxygen and produce H^O^- and also phagocytose under anoxic conditions. Luminol-dependent chemiluminescencc, well-known for this ani- mal, can be eliminated by N, purging, indicating a direct oxygen dependence and that stored, metastable reactants are not involved. The presence of a hydroperoxide metabolism in a diffusion dependent animal with hypwxic tissues raises questions about the utility of H^O-, production. It is suggested that oxygen-denved molecules detected by luminol-dependent cheniiluminescence may be a consequence and measure of a routine, activity related, oxygen metabolism. EMERGING EVIDENCE OF EXTRACELLULAR PRO- TEASES AS IMPORTANT VIRULENCE FACTORS OF PERKINSUS MARINUS. Jerome F. La Peyre,* Kathleen A. Garreis, Heather A. Yarnall, and Mohamed Faisal, Virginia Institute of Marine Science. College of William and Mary. Gloucester Point. VA 23062. Our recent studies have shown that Perkinsus marinus secreted multiple serine proteases. Current investigations are being per- formed to determine the role these proteases may play in the sur- vival and pathogenicity of this deadly protozoan. Serine proteases constitute the major part of the protozoan's extracellular proteins and are secreted by all P. marinus isolates so far examined. At physiological pH. P. marinus proteases (PMP) hydrolyzed a wide range of proteins including extracellular matrix proteins (e.g. fibronectin). PMP appeared essential for the devel- opment of P. marinus since protease inhibitors suppressed the parasite propagation. Recent findings have also shown that hemocyte motility is reduced in a dose dependent manner by exposure to PMP. More- over, lysozyme activities and hemagglutinin titers of oyster plasma were suppressed as a result of coincubation with purified PMP. Investigations done in vivo suggested that PMP may be impor- tant for these establishment of infection and propagation of the parasite. For example, we have recently found that the parasite body burden in oysters administered liposomes containing P. marinus extracellular proteins and then challenged with P. mari- nus. increased significantly compared to oysters fed liposomes containing only seawater. There is thus increasing evidence suggesting that PMP play a role in the invasion and growth of the parasite in host tissue and also in counteracting both cellular and humoral host defenses of the oyster. MOLLUSCAN FEEDING STUDIES KINETICS OF DIARRHETIC SHELLFISH TOXINS IN THE BAY SCALLOP, ARGOPECTEN IRRADIANS. Andrew G. Bauder* and Jon Grant, Department of Oceanography. Dal- housie University, Halifax, NS, Canada, B3H 4J1: Allan D. Cembella and Michael A. Quilliam, Institute for Marine Bio- sciences, National Research Council, Halifax, NS, B3H 3Z1, Diarrhetic shellfish poisoning (DSP) is a worldwide public health risk which also constitutes an economic threat to the com- mercial harvest of shellfish. Bivalve molluscs can acquire DSP toxins by ingesting toxic dinoflagellates from the water column and from benthic seston. Although there have been field studies relating the incidence of DSP toxins in shellfish to toxic di- notlagellate blooms, few studies have described DSP toxin kinet- ics in bivalves under either natural or controlled laboratory con- ditions. In the present study, the dynamics of DSP toxins were exam- ined in juvenile and adult bay scallops by feeding cells of the epibenthic dinotlagellate Prorocentrum lima, a known producer of DSP toxins, to scallops in controlled laboratory microcosms. Liq- uid-chromatography combined with ion-spray mass spectrometry (LC-MS), a powerful new analytical method for marine toxin de- tection, was used to analyze for DSP toxins in dinotlagellate cells and scallop tissues. Toxin kinetic parameters, including rates of toxin uptake, biotransformation, anatomical compartmentalization and detoxification were determined. The physiological effects of 502 Abstract . 1996 Annual Meeting. April 14^18, 1996 National Shellfisheries Association. Baltimore. Maryland toxic dinoflagellates on scallop feeding activities and survival were also established. Juvenile and adult clearance rates were not inhibited by expo- sure to toxigenic P. lima cells and no scallop mortalities were observed. Relatively high weight-specific ingestion rates demon- strated that scallops could exceed regulatory toxin limits (0.2 (jig DSP toxin g ^ ' wet wt. ) in less than one hour of exposure to high P. lima cell densities. Toxin saturation levels (2 ^.g gi wet wt.) were attained within the first two days of exposure, however toxin retention efficiency was very low (<5%). Although most of the total toxin body burden was associated with visceral tissue, weight-specific toxin levels were also high in gonads of adult scallops. Rapid toxin loss from gonads within the first two days of depuration indicated that the toxin was derived primarily from a labile (unbound) component within the intestinal loop section through the gonads. Detoxification of visceral tissue, however, followed a biphasic pattern of rapid toxin release within the first two days of depuration, followed by a more gradual toxin loss over a two week period, suggesting that fecal deposition may be an important mechanism for rapid release of unassimilated toxin and intact dinotlagellate cells. MUCOCYTE DISTRIBUTION AND RELATIONSHIP TO PARTICLE TRANSPORT ON THE PSEUDOLAMELLI- BRANCH GILL OF CRASSOSTREA MRGINICA (BI- VALVIA: OSTREIDAE). Peter G. Beninger* and Suzanne C. Dufour, Departement de Biologic. Universite de Moncton. Monc- ton. N.B. Canada ElA 3E9. To gain a more complete understanding of particle transport on the bivalve pseudolamellibranch gill, the mucocyte secretion types and distribution were determined for this organ in the East- em oyster Crassostrea virginica, and related to previous endo- scopic data. Three adult oysters were collected from Shediac Bay (New Brunswick. Canada) in July 1994. immediately fixed, then dissected and processed for histology and whole mount mucocyte mapping using a modification of the periodic acid — Schiff — alcian blue protocol. One type of mucocyte contained acid-secretion mucopolysaccharides (AMPS), while the other type contained neutral mucopolysaccharides (NMPS). A clear gradient in mucocyte density was observed from the plical crest to the trough; for all but the anteriormost 15 plicae the proportions of each mucocyte type remained constant; the 15 anteriormost plicae presented an increased proportion of AMPS. These propor- tions would produce a relatively viscous acid-dominant mucus after mixing on the ciliated gill surface. The principal filament troughs contained relatively few mucocytes, aligned on the median ridge. This arrangement could account for the observed range of particle velocities in the.se filaments. The results of the present study conform to a pattern of specialization of mucus types and functions on the diverse types of bivalve gill, depending on par- ticle trajectory, transporting surface architecture, and dominant current flow. DIFFERENTIAL SENSITIVITY AND PSP TOXIN ACCU- MULATION IN TWO CLAM SPECIES, SPISULA SOLIDIS- SIMA AND MYA ARE^ARIA. Monica V. Bricelj* and David Laby, Marine Sciences Research Center. State University of New York. Stony Brook, NY 11794-5000; Allan D. Cembella, insti- tute for Marine Biosciences, National Research Council, 1411 Oxford Street. Halifax, NS B3H 3Z1. Canada. Differences among bivalve species in the rate of accumula- tion of paralytic shellfish poisoning (PSP) toxins are generally correlated with their in vitro nerve sensitivity to saxitoxin (STX) and ability to maintain active feeding during toxic dinoflagel- late blooms. In this study, the relative sensitivity of juvenile sur- fclams, Spisula solidissima, and softshell clams, Mya arenaria, to PSP toxins was determined from their burrowing activity in sed- iment, and siphon responsiveness (retraction). These two behav- ioral components are likely to influence survival of natural popu- lations. Long-term (lid) exposure to the high-toxicity dinoflagel- XaXt Alexandrium tamarense icf. excavatum) (clone PR 18b) had no effect on burrowing or siphon retraction in Spisula. but both be- haviors were severely impaired in Mya, within a few hours of exposure to toxic cells. Although overall, both behavioral re- sponses showed similar temporal patterns, burrowing was more severely affected and provided a more reliable and consistent sen- sitivity index than siphon tlaccidity. The Mya test population, which had no prior history of exposure to PSP toxins, was char- acterized by high individual variability in toxin sensitivity: ca. 27% of clams tested repeatedly during toxification generally showed no burrowing inhibition. This provides the first evidence of intrapopulation variability in sensitivity to PSP toxins in a bi- valve species. Both species accumulated most of their body burden of toxin (82 to 89%) in viscera during toxin uptake, but Spisula showed a 23-fold higher rate of toxin accumulation and maximum toxicity than Mya under identical experimental conditions. The toxin com- position of ingested dinoflagellates exhibited no significant changes in Myo tissues. In contrast, Spisula showed a high capac- ity for in vivo metabolic transformation of individual PSP toxins, as evidenced by the production of potent decarbamoyl gonyautox- ins (dcGTXs) from weakly potent N-sulfocarbamoyl (C-) toxins, as well as gonyautoxins, present in ingested cells. Enzymatic con- version of PSP toxins to decarbamoyl derivatives has been previ- ously confirmed in only three Pacific clam species. This study identifies likely pathways for de novo production of these toxins (dcGTXs and dcSTX) in Spisula. and discusses the implications of our laboratory results in explaining the differential retention of PSP toxins in field populations of these two bivalve species. National Shellfishcrics Association, Baltimore, Maryland Abstract. IW6 Annual Meeting, April 14-18, 19% 503 TEMPORAL PERSPECTIVES ON FEEDING AND DIGES- TION BY SUSPENSION-FEEDING BIVALVES. Peter J. Cranford* and Barry T. Hargrave, Department of Fisheries and Oceans, Bedford Institute of Oceanography, P O Box 1006. Dart- mouth, Nova Scotia B2Y 4A2; Conrad A. Pilditch, Department of Oceanography. Dalhousie University. Halifax. N.S. B3H 4J1. Food acquisition by bivalves in the natural environment is driven by a complex interplay between the different time scales of variation in oceanographic variables, time constraints on acclima- tion capabilities and temporal variations in energy demands. To provide empirical data to help predict temporal changes in bivalve food acquisition, sea scallop iPlacopecten magellaniciis) and mus- sel {Mytiliis edulis) feeding behaviour was monitored in situ over time-scales ranging from hours to months. Food acquisition was monitored at several coastal sites in Nova Scotia using a new sediment trap method that provides time-series data on the ingestion, absorption and egestion rates of bivalve populations over hourly to daily intervals. Replicate trap data dem- onstrate the high precision of the method. Short-term feeding re- sponses (integrated over hourly intervals) of scallops to environ- mental changes show that periodic feeding activity can account for a high proportion of daily food intake and that food quality (or- ganic content of seston) explains (93% of the variation in absorp- tion efficiency. Seasonal variations in scallop and mussel feeding responses (integrated over daily intervals) were monitored concur- rently over four 40 day sampling periods during the spring, sum- mer, and fall of 1995. Data are currently being analyzed and results will be discussed in the context of inter-specific variations and the relative importance of possible exogenous forcing func- tions (food supply, flow, and temperature) and endogenous con- trols (energy demands and acclimation capabilities) on bivalve feeding physiology. FIELD AND MODELLING STUDIES OF BIVALVE CUL- TURE IN A BOREAL ENVIRONMENT. Jon Grant* and Conrad A. Pilditch, Department of Oceanography. Dalhousie University. Halifax. Nova Scotia, Canada BON 2S0. In northern climates, the spring phytoplankton bloom inevi- tably occurs during periods of low water temperatures. Based on temperature functions of feeding, bivalves cultured under these conditions (e.g. Myliltis eiinlis in suspended culture) thus face possible temperature-inhibition of ingestion during the spring bloom. Field studies of particulate food supplies and mussel growth as well as field estimates of feeding rates indicate that (1) primary production is inhibited by ice cover. (2) winter food supplies have low organic content despite high concentrations. (3) mussels display reduced absorption efficiency with this food supply. (4) and mussels have a negative energy balance during winter. Because mussels are in poor condition at the end of win- ter, the spring bloom is essential in promoting summer growth. Simulation modelling of mussel scope for growth in early spring suggests that acclimation of feeding to low temperatures allows animals to take advantage of the spring bloom, with accelerated growth during this period. The timing of the spring bloom is also critical to the growth trajectory at this time. Further implica- tions of food and temperature limitation for bivalve growth are discussed. DELIVERY OF LOW MOLECULAR WEIGHT, WATER- SOLUBLE NUTRIENTS TO MARINE SUSPENSION FEED- ERS. Chris J. Langdon* and Mike Buchal, Hatfield Marine Science Center and Department of Fisheries and Wildlife. Oregon State University. Newport. Oregon 97365. A major problem in formulating artificial diets for manne sus- pension-feeders is the development of microparticle types that will effectively deliver low molecular weight, water-soluble nutrients to these organisms. Retention of nutrients, such as amino acids, by alginate, carrageenan or zein microgel particles is poor, with more than 80% being lost within two minutes after suspension in sea- water. Lipid-walled microcapsules retain amino acids and other water- soluble, low molecular weight nutrients with greater efficiency than microgel particles. Lipid-walled microcapsules with walls prepared with tripalmitin retain more than 90% riboflavin after 24 h suspension in seawater. Feeding experiments indicate that shrimp (Panaeus vannamei) mysids are capable of breaking down tripalmitin capsule walls and liberating capsule contents for assim- ilation. Unfortunately, we found that tripalmitin-walled capsules were not efficiently digested in feeding experiments with clams (Tapes philippinarum). Softening the capsule walls, by inclusion of 40% w/w of the triacylglyceride fraction of fish oil. improved capsule digestibility but reduced retention of encapsulated nutrients; for example, after 24 h suspension is seawater, lipid-walled capsules with walls composed of 60% w/w tripalmitin and 40% w/w fish oil retained only 21% of encapsulated oxytetracycline. Clearly, opti- mal microparticles for the delivery of encapsulated nutrients to suspension-feeders should show both efficient retention and di- gestibility of encapsulated material. FEEDING ACTIVITY IN THE SEA SCALLOP PLA- COPECTEN MAGELLAN ICUS: COMPARISON OF FIELD AND LABORATORY DATA. Bruce A. MacDonald*, Uni versity of New Brunswick. Saint John. NB. E2L 4L5 Canada; J. Evan Ward, Department of Biological Sciences. Salisbury State University. Salisbury, MD 21801; Gregory S. Bacon, Department of Fisheries and Oceans. Moncton, NB. EIC 9B6 Canada. Suspension-feeding bivalves are exposed to a food supply that fluctuates unpredictably in both concentration and quality of the particles. Many studies have attempted to understand how these 504 Abstract. 1996 Annual Meeting, April 14-18, 1996 National Shellfisheries Association, Baltimore, Maryland bivalves are adapted to exploit their food resource, through feed- ing processes such as capture, selection and absorption, and main- tain the flow of energy into growth and reproduction. The objec- tive of this study was to measure the variety of feeding responses sea scallops [Placopecten mageltanicus) exhibit over the range of environmental conditions they encounter in their natural environ- ment. To accomplish this we exposed scallops to a complete range of temperature conditions and the natural assemblage of particles by mooring a small research vessel at various experimental sites throughout the year. To complement the field work we also stud- ied feeding responses in the laboratory where the concentration ( 1 , 3, 7 and 14 mg I" ') and quality (25, 50, and 80% organics) of the diet were manipulated, by varying the proportions of organic and inorganic components, to simulate field conditions. Clearance rates typically decline with increases in seston concentration but appear to be independent of seston organics and ambient temper- atures (0-15°C). Absorption efficiency increases proportionately with increasing seston organics. is independent of seston concen- tration and appears, at least initially, to be negatively related to increasing ambient temperatures. There was often close agreement between results from the field and the laboratory suggesting that fairly realistic feeding rates may be obtained using controlled diets in the laboratory. IN SITU MEASUREMENTS OF MUSSEL [MYTILUS EDVLIS) ENERGY ACQUISITION IN RELATION TO SESTON CONCENTRATION IN A SUBTIDAL MAINE ES- TUARY: HOW IMPORTANT IS THE SHELL GAPE RE- SPONSE? Carter R. Newell.* Great Eastern Mussel Farms, Inc., Tenants Harbor, ME 04860, and Department of Biology, University of New Brunswick, St. John. New Brunswick, Canada E2L 4L5. Mussel scope for growth was estimated using water pumped 100 m to a flow-through apparatus where filtration rates and res- piration rates were measured during low, flood, high and ebb stages of the tide. Repeated measures ANOVA indicated signifi- cant differences in filtration rates (p > 0.001) between tidal stages. Time-lapse video records of pumping rate estimated by shell gape (as % maximum per individual) in situ showed a similar pattern. A sustained 2-3 hour period of reduced gape during low ambient food was observed around low tide. This was accompa- nied with a drop in oxygen consumption. Under ambient conditions of 1-25 million particles per liter, filtration rate was significantly positively correlated with concen- tration at low concentrations, and was significantly negatively cor- related with concentration at high concentrations. Control of pumping rate using the shell gape response is a mechanism by which subtidal mussels couple maximum energy gain with maximum seston supply. Daily filtration rate is not siinply a response to average daily conditions, but rather repre- sents a dynamic response to a fluctuating seston regime. SESTON SUPPLY TO SCALLOPS IN SUSPENDED CUL- TURE. C. A. Pilditch* and J. Grant, Oceanography Depart- ment. Dalhousie University, Halifax N.S., Canada B3H 4J1: A. L. Mallet and C. E. A. Carver, Department of Fisheries and Oceans, Biological Sciences Branch, PC Box 550 Halifax N.S., Canada B3J 2S7; P. J. Cranford, Department of Fisheries and Oceans. Habitat Ecology Division, Bedford Institute of Oceanog- raphy, PO Box 1006. Dartmouth N.S., Canada B2Y 4A2. The filtration activity of dense aggregations of bivalves can locally deplete the water of seston resulting in food limited growth. Under these conditions, the currents that supply seston cannot offset seston removal due to feeding. We conducted a field and modelling study that examined the tidally driven supply of seston to a culture of suspended scallops (Placopecten mageltan- icus) in Whitehaven Harbour, Nova Scotia. The goal of the re- search was to predict optimal stocking densities and lease geom- etries which minimize food limitation. Measurements of seston concentration, diet quality and current velocity were made over six consecutive tidal cycles inside and outside the lease. Scallop ingestion rates were estimated bihourly from biodeposition rates and changed in response to environmental forcing. Current velocities in the centre of the lease were 50% lower than those measured at the reference site. Despite the re- duction in flow, seston concentrations inside the lease were not significantly lower than those measured outside the lease. Results suggest that at present stocking densities food limitation did not occur. The field data were used to parameterize a scallop feeding sink term in a one dimensional advection-diffusion mod- el of changes in seston concentration in a lease. The model is used to suggest a optimal scallop density for the Whitehaven lease. The possibility of extending this model to two dimensions to show the effect of lease orientation relative to the predominant flow direction will be discussed. Modelling results suggest that an in- crease in the size of the lease, and/or stocking density, could reduce the supply of seston to growth limiting levels in the centre of the lease. SEDIMENTING PHYTOPLANKTON AS A MAJOR FOOD SOURCE FOR SUSPENSION-FEEDING QUEEN SCAL- LOPS (AEQUIPECTEN OPERCULARIS L.) OFF ROSCOFF (WESTERN ENGLISH CHANNEL)? Gerard Thouzeau,* Frederic Jean, and Yolanda Del Amo, URA 1513 CNRS, UFR Sciences et Techniques, 6 avenue Le Gorgeu. B.P. 809, 29285 BREST Cedex France. The chlorophyll and phaeopigment content of bottom-water, sediment and gut o{ Aequipecten opercularis (Mollusca, Bivalvia) was measured together with seasonal changes in weight and re- production of the pectinid, offshore Roscoff (western English Channel). The study site (site no. 1 of the PNOC; 48°51'00 N, 3°54'00 W; 71 m depth) belongs to the sand-gravel Venus fasciata community which extends on clean coarse sediments of the west- em English Channel, in areas generally deeper than 65-70 m. National Shellfisheries Association. Baltimore, Maryland Abstract. 1996 Annual Meeting. April 14-18, 1996 505 There is no vertical stratification of the water column in summer (mixed waters); water temperature ranges from 9°C in February to 15.4°C in August. Chlorophyll concentration in surface and bot- tom waters is typically 2.5-3.0 (xg • I ' during spring bloom (May-June), vs. less than 1.0 jjig • 1 ' otherwise. Sediment or- ganic matter content ranged from 2.5 to 6.4'^ in 1992 and 1993. Seasonal variations of sediment organic matter content were cor- related with phytoplankton sedimenting events, emphasizing pe- lagic-benthic coupling. High seasonal variations were observed in the general condition of a {(standard)) 50 mm shell length A. opercularis: in 1992 and 1993, gonad dry weight, body dry weight and stomach pigment content showed a sixteen-fold. three-fold and ten-fold variation, respectively. Sedimentation of the spring phytoplankton bloom as the main regulating factor for weight in- creases and development of reproductive tissue is questioned since variations of pigment uptake may not correlate with variations of pelagic trophic inputs to the benthos. Chlorophyll and phacophytin concentrations in bottom-water were highest from April to June, while stomach pigment content was lowest (maximum values in August and September). In contrast, sedimentation of summer phytoplankton blooms would be a main regulating factor for body weight increase. WHEN IS IT TIME TO FEED THE SCALLOPS? G. H. Wik- fors, B. C. Smith, J. H. Alix, and M. S. Dixon, NOAA, Na- tional Marine Fisheries Service. Northeast Fisheries Science Cen- ter, Milford, CT 06460: M. S. Dixon, Marine Sciences and Tech- nology Center. University of Connecticut, Groton, CT 06340. As part of a research program to develop biochemically-based feeding standards for post-set bay scallops. Argopecten irradians. we conducted an experiment to determine how often to feed post- set scallops to achieve maximal growth and feed conversion effi- ciency. To accomplish this expenment. we designed and built a PC- controlled fluidics system for 12 molluscan rearing chambers that permits unattended control of dis-continuous feeding in each cham- ber. Experimental design compared growth of 5 mm scallops on a unialgai diet of Tetmselmis chiii. strain PLY429. fed every 24, 12,6, or 3 hours. Daily ration and cumulative feeding time were identical for all experimental feeding regimes. Scallops fed every 6 hours grew more rapidly than those fed more or less often. The optimal 6-hour interval corresponds with the 6-hour tidal regime in the scallop's estuarine habitat, leading us to speculate that the scallop's digestive cycle is adapted to process food on this schedule. This finding suggests that knowl- edge of feeding behavior in nature may provide guidelines for farming bivalve mollusks. MOLLUSCAN NUTRITION FEEDING BEHAVIOUR AT HIGH AND VARIABLE SES- TON LOADS. Brian Bayne* and Tony Hawkins, Plymouth Ma- rine Laboratory, Prospect Place, Plymouth PLl 3DH, Devon, UK. Suspension-feeding bivalves live in a diversity of food envi- ronments in which the suspended particulate matter may vary in concentration from low (<2 mg dry mass 1~ ') to extremely high ( > 1 00 mg 1 " ' ) values; the nutritional quality of this material may also differ, from diets high in living phytoplankton to those dom- inated by resuspended silt of low nutritious value. Both quantity and quality of food may vary over timescales from minutes, through tidal (hours and days) to seasonal scales. Such are some of the extrinsic ( = supply) variables with which the animals must cop)e. In addition, there are intrinsic ( = demand) variables, which result in individual and species variability and add further complexity to the analysis of feeding behaviour. Feeding may be considered as a linked sequence of traits which includes: filtration of material from suspension: sorting of this material into particles that are rejected and those ingested; digestion, absorption and assimilation of nutrients from the ingesta; egestion of non-absorbed matter. We need to understand the rates and efficiencies with which each of these processes is effected under different conditions of nutnent supply. An additional challenge is to understand the causes of intrinsic vanability. both between individ- uals and between species. In our paper we will review recent studies on various bivalve species whose objectives have been to analyse feeding behaviour under environmentally realistic conditions. We will seek general- ity for the processes of suspension-feeding itself, across a spec- trum of food environments and of species. And we will briefly address questions of individual and species variability that pose particular challenges for the future. OBLIGATE ENDOSYMBIOTIC ASSOCIATIONS BE- TWEEN CHEMOAUTOTROPHIC BACTERIA AND MA- RINE BIVALVES: NUTRITIONAL IMPLICATIONS TO THE EARLY LIFE STAGES. Craig S. Cary. College of Ma rinc Studies, University of Delaware, Lewes, DE 19958. The occurrence of obligate symbiotic associations between bacteria and invertebrate hosts appears widespread in marine or- ganisms. These unique symbiotic associations are particularly abundant in marine mollusks where they are now know to occur in 5 families of bivalves and 2 other orders of mollusks. Whether the bacteria reside externally (episymbiosis) or within specialized cells of the host gill tissue (endosymbiosis) it is clear than in most circumstances host development and settlement are constrained by the initial acquisition and specific energy requirements of the sym- biont. Biological communities associated with deep-sea hydrother- mal vents and other reducing environments inhabit a highly vari- able and ephemeral environment characterized by large fluxes in abiotic and biotic conditions. Consequently, the reproductive, dis- persal and nutritional strategies of the resident fauna must often accommodate relatively narrow windows of opportunity. In the bivalves, this appears to be achieved through continuous and broad range dispersal capabilities coupled with the vertical transmission 506 Abstract. 1996 Annual Meeting, April 14-18. 1996 National Shellfisheries Association. Baltimore. Maryland of the symbionts. These chemoautotrophic symbioses range from obligatory relationships in which the adult hosts completely lack digestive tracts to those which maintain fully functional particulate feeding in adult life. Still others are thought to have a functional feeding apparatus durmg the early life stages which is lost in the adult form. In all cases the distribution of the host is clearly related to factors directly associated with its symbiont's energy require- ments (i.e. where both reduced sulfur or methane and molecular oxygen co-occur). It is therefore conceivable that the symbionts play an active role in habitat selection and the induction of meta- morphosis and settlement of the host. Although much of the early life history of the vent organisms and their shallower water analogs has remain unresolved molecu- lar techniques are providing some clues as to the importance of the symbiont in the early life history of the host. Nucleic acid probing technologies provide the resolution necessary to identify sym- bionts even within the embryos providing key information of their role in early development. LIPID NUTRITION AND FATTY ACID SYNTHESIS IN OYSTERS. Fu-Lin E. Chu, Virgmia Institute of Marine Sci- ence, School of Marine Science, College of William and Mary. Gloucester Point, VA 23062. Lipids play a unique role in reproduction of oysters. Egg stor- age lipids are important for early larval development. A fatty acid metabolism study revealed that de novo synthesis of saturated fatty acids (C16;0 & C18:0) occurred in adult oysters with reutilization of '''C-acyl groups derived from (i-oxidation. However, only lim- ited elongation and no desaturation of '''C-labeled palmitic, lin- oleic and linolenic acids (C16;0, C18;2-n6, C18:3-n3) was ob- served. Similarly, <1.0'7f of dietary '■'C-18 precursors were elon- gated to eicosapentaenoic (EPA, C20:5-n3) and docosahexaenoic (DHA, C22:6-n3) acids in spat. For reproductive success, EPA, DHA, CI8:2-n6 and C18:3-n3 have to come mainly from a dietary source. Phytoplankton is the major food source of bivalve mol- luscs and is considered to be the main supplier of C18:2-n6. CI 8; 3-n3, EPA and DHA in the marine food web. During phytoplank- ton blooms and whenever food is available, oysters store glyco- gen, rather than lipid as an energy reserve. They actively assimilate dietary fatty acids into their lipids prior to or during oogenesis. The seasonal variation of omega-3 PUFAs in the vis- ceral mass corresponded to seasonal changes in the diet. There is evidence of active assimilation/incorporation of dietary fatty acids into the neutral lipids and phospholipids of oysters. During oo- genesis, oysters probably synthesize lipids using recently ingested and stored fatty acids and at the expense of stored glycogen. Sea- sonal variation of lipids and omega-3 polyunsaturated fatty acids ■PUFAs) in the visceral mass were related to the oyster's repro- ductive cycle. Total lipid and omega-3 PUFAs in the visceral mass wci.: higher prior to the spawning season than after spawning. CILIARY SUSPENSION-FEEDING AND PARTICLE SE- LECTION IN MOLLUSC LARVAE. S. M. Gallager. Woods Hole Oceanographic Institution, Woods Hole, MA 02543. Most planktonic Mollusc larvae use cilia for feeding and loco- motion. A brief review of the mechanisms available to Mollusc larvae for capturing particles and propelling fluid is presented together with appropriate morphological and functional constraints placed on each process. Mollusc larvae are used as models to illustrate how more than one capture mechanism may be function- ing simultaneously depending on particle size and surface pro- perties. For example, bivalve larvae capture flagellates at the base of the cilia using hydrodynamic retention, a mechanism for enhancing direct interception between cilia and non-sticky particles. Diatoms, however, are captured at the tips of the cilia by direct interception. Adhesion between cells and the tips of the cilia is enhanced by the cells" sticky surface which changes during cell growth. Particle encounter, capture, trans- port to the mouth, and selection for ingestion have distinct pro- babilities which must be observed and quantified indepen- dently and under a wide variety of environmental conditions to obtain accurate predictions of feeding success in suspension- feeders. OMNIVORY BY THE MUSSEL, GEUKENSIA DEMISSA. Daniel A. Kreeger.* Patrick Center for Environmental Research, Academy of Natural Sciences, Philadelphia, PA 19103; Roger I. E. Newell. Horn Point Environmental Laboratory, University of Mar>'land, Cambridge, MD 21613. Suspension-feeding bivalves must derive their nutrition from a complex suite of natural microparticulate material when phyto- plankton are scarce. This is especially true for the ribbed mussel, Ceukensia deinissa. which is a prominent inhabitant of the high intertidal zone of eastern USA salt marshes. Ribbed mussels have numerous physiological adaptations for the acquisition and assim- ilation of alternative food particles, including vascular plant detri- tus, bacteria (free, aggregated and attached), benthic diatoms, and heterotrophic flagellates. As part of our ongoing program to de- termine how this mussel meets it's carbon and nitrogen require- ments, we recently measured it's ability to ingest and digest cel- lulolytic bacteria and various identified species of heterotrophic flagellates and benthic diatoms. Ribbed mussels ingested radiola- beled carbon from bacteria, heterotrophic flagellates and benthic diatoms with efficiencies that were 59%, 88% and 78%, respec- tively, compared with a their carbon ingestion rates for a reference species of phytoplankton, hochrysis galbana (clone T-ISO). Fur- thermore, carbon was assimilated from bacteria, heterotrophic flagellates and benthic diatoms with efficiencies (42, 44% and 87%, respectively) that were comparable to that from T-ISO {11%). We are currently measuring seasonal variation in the abundance of these alternative foodstuffs in situ to estimate their relative contributions to the mussel's carbon and nitro- gen demand on an annual basis. Nevertheless, in comparison to National Shellfisheries Association, Baltimore, Maryland Absircul. 1996 Annual Meeting. April 14-18. 1996 507 other suspension-feeding bivalves, G. demissa clearly has an ex- ceptional capacity for omnivory. which could be a critical nutri- tional adaptation for life in a detritus-rich, phytoplankton-poor habitat. HOW ARE BIVALVE BROODSTOCK AND LARVAE ADAPTED TO MEET THEIR NUTRITIONAL REQUIRE- MENTS FOR LIPIDS. Philippe Soudant, Yanic Marty, Jean Rene Le Coz, Jeanne Moal. and Jean Francois Samain,* In- stitut Francais pour I'Exploitation de la mer (IFREMER). Centre de Brest. BP 70, 29280. Plouzane. France. Three microalgal diets, differing in composition in the levels of 20:4(n-6). 20:.'i(n-3) and 22:6(n-3) fatty acids and in sterol con- tent, were used to study the nutritional lipid requirements for re- production and larval development of the scallop Peclen imuimtis. A mixture of Isochnsis galhcina (clone T. Isol (T) and Chaelo- ceros cakitniiis (C) plus Pavknu liillteri (P) and SkeleUmema costatum (S) was fed to broodstock. and a mixed diet (PTS) was fed to larvae. Separation of the different polar lipids was per- formed using HPLC, then GC for their fatty acid composition. Phospatidyl choline (PC) and plasmalogens (PLSM) were the ma- jor classes identified, phosphatidyl inositol (PI), phosphatidyl serine (PS), phosphatidyl ethanolamine (PE). and a glycolipid (GLY) were minor classes. Peclen »u;.v™i« appears to regulate the composition of its polar lipid and sterol classes under different dietary conditions. In the PC class. 20;5(n-3) was not accumulated, but 22:6(n-3) and 20; 4(n-6) were selectively accumulated compared to levels in the diet. The same fatty acids were selectively retained in plasmalogens when their concentration in the food was low. and levels of 22; 6(n-3) and 20;5(n-3) differed in oocyte and sperm. Lastly, high selectivity was observed for 20;4(n-6) in PI. 22;6(n-3) in GLY. and 20;5(n-3) in PE. resulting in nearly constant levels that were almost independent of food composition. In the same way. cho- lesterol was preferentially concentrated, compared to other phy- tosterols. when the diet was deficient in this sterol. These differ- ences in fatty acid, sterol or lipid class selectivities arc discussed in reference to known or putative requirements for reproduction and larval development. LOOKING INTO THE "BLACK BOX": FEEDING STRA- TEGIES AND LIMITATIONS OF SUSPENSION-FEEDING BIVALVES. Evan J. Ward,* Department of Biological Sci- ences. Salisbury State University, Salisbury. MD 21801; Jeffrey Levinton, Department of Ecology & Evolution. S.U.N.Y.. Stony Brook, NY.. 11794; Sandra Shumway, Natural Science Divi- sion. Southampton College. Southampton. NY 11968; Terri Gucci, Bigelow Laboratory for Ocean Sciences. Boothbay Har- bor. ME 04575. Benthic particle feeders are exposed to a food supply that varies in quality and quantity along both spatial and temporal scales. Previous studies have shown that bivalves deal with such fluctu- ating particle regimes in a variety of ways, including adjustments in pumping and ingestion rates, and rejection of non-nutritive par- ticles as pseudofeces. Data such as these have led researchers to develop and test models of feeding compensation for changes in particle supply, but most models treat pallial organ processes like a "black box" with an input (capture/collection), one branch (pseudofeces) and an output (ingestion). There is evidence, how- ever, that significant adjustments occur at a much finer scale. To demonstrate some of these fine scale adjustments, we ex- posed the oysters Crassoslrea virginica and C. gigcis to a mixture of ground, aged Spartina sp. (3-10 |j.ni) and similar sized phyto- plankton (Rhodomonas sp. or Teiraselmis sp.) at three concentra- tions (I0\ lO''. and 10"^ particles ml" '). We then examined the gills and labial palps by means of endoscopy and sampled, in vivo, particulate material from various ciliated tracts. These samples were then analyzed with a flow cytometer. Preliminary results indicate that in oysters the gill is the main sorting organ, whereas the labial palps function more for controlling the volume of ma- terial to be ingested. Our study is the first to take a truly functional approach in modeling the way in which pallial structures respond to changing particle fields. Studies such as these will lead to a better understanding of pallial organ function, and help define the influence of bivalves on trophic dynamics of benthic ecosystems. MOLLUSCAN REPRODUCTION INDUCTION OF TRIPLOIDV IN UNCONDITIONED EASTERN OYSTERS, CRASSOSTREA VIRGINICA, USING NITROUS OXIDE UNDER INCREASED PRESSURE. James W. Anderson* and Richard K. Wallace, Auburn University Ma- rine Extension and Research Center. 4170 Commander's Drive. Mobile. AL 36615. Triploidy is commonly induced using cytochalasin B in Pacific oysters, Crassostrea gigas, to increase production. Use of cyto- chalasin B has raised safety concerns and there is interest in de- veloping alternate methods for producing triploid oysters. Eastern oysters. Crassostrea virginica. were collected from Mobile Bay, Alabama and spawned by manually extracting the gametes. Eggs were treated with nitrous oxide at several pressures ranging from 0 to 19 atm. Initiation time and duration depended on development of the fertilized eggs. Triploidy was induced in up to 56. 87^ of the larvae, using a 15 min nitrous oxide treatment at 7 atm. When nitrous oxide was applied at ambient pressure, there was a maxi- mum triploidy at 8.3%. Effects of treating eggs for different durations were examined. Triploidy in the 20 min duration group (38.7%) was significantly different from the 10 min duration (16.0%) and control (6.0%) groups. Average survival rates to straight-hinge were 5. 22, and 52% for the 20 min duration, 10 min duration, and control groups, respectively. Nitrous oxide under pressure can induce triploidy in Eastern oysters when treatment conditions are ideal. 508 Abstract. 1996 Annual Meeting, April 14-18, 1996 National Shellfisheines Association, Baltimore. Maryland RECRUITMENT PATTERNS OF MYA ARENARIA L. FROM EASTERN AND SOUTHWESTERN MAINE: II. EF- FECTS OF SITE, TIDAL HEIGHT, AND PREDATOR EX- CLUSION. Brian F. Beal* and K. W. Vencile, University of Maine at Machias. 9 O'Brien Avenue, Machias. ME 04654; Stephen R. Fegley, Maine Maritime Academy, Castine. ME 04421. Soft-shell clam landings in Main declined 67% from 450.000 bushels harvested in 1983 to less than 150.000 in 1994. The de- cline was spatially variable along the coast as landings in eastern Maine, where, historically 45-65% of all clams are harvested annually, dropped by 90% during this eleven-year period, but increased a modest 15% over the same time in southwestern por- tions of the coast. Reasons for the decline were not attributable to differences in harvesting pressure between the two coastal regions. We invoke three competing hypotheses to explain dif- ferences in relative clam abundances between the two regions: 1) differences in fecundity. 2) differences in larval survival, and 3) differences in post-settlement abundance and/or mortality. We employed a full-factorial, completely randomized block design at six intertidal sites in eastern Maine and six sites in southwestern Maine to examine the effects of substrate type (mud vs. sand), tidal height (low vs. high), and predator exclusion (protected vs. unprotected) on post-settlement survival success. If no apparent differences in post-settlement abundance/ mortality exist between regions, two hypotheses remain to be tested. Beginning in late March and continuing through April. 1995. we initiated a "long-term"' experiment to examine the cumula- tive effects of settlement processes at each of twelve sites along the Maine coast. At each site. 120 plastic flower pots (II cm diameter x 9.3 cm deep) filled with terrestrial "mason" sand (mean 67% decrease) while annual landings in southwestern Maine have expanded to over 50.000 bushels (~ 15% increase). Reasons for the changes in landings cannot be attributed to regional differ- ences in cither relative harvest pressure or to clamflat closures: lower abundances of soft-shell clams currently exist in eastern Maine where, historically 45-65% of all clams harvested in Maine were taken. Anecdotal and direct observations of intertidal flats indicated that eastern Maine soft-shell clam populations experi- enced repeated recruitment failures in the past decade while south- western clam populations displayed fairly consistent annual re- cruitment. We questioned whether the absence of visible clam spat in eastern Maine occurred because of a failure of soft-shell clam larvae to reach the flats or because of high mortality of recently settled juveniles. To distinguish between these alternatives we employed a full-factorial, completely randomized block design that varied site (two sites in eastern Maine versus two sites in southwestern Maine), tidal height (mid versus low), and predator exclusion (protected versus unprotected). In mid April we placed, at each tidal height at each site, 80 plastic flower pots ( 1 1 cm diameter x 9.3 cm high) filled with "mason" sand (mean i^ diameter = 2) arrayed into 20 replicate blocks of four pots each in a 2 x 2 matrix. The pots were buried flush to the sediment surface and two pots within each block were covered with flexible, black plastic netting (aperature = 6.4 mm) to exclude large predators. At approximately two week intervals through August (and biweekly or monthly, depending on site, until early November) all pots at all sites were replaced with new pots filled with mason sand. The exposed pots were returned to the laboratory. The top 1 cm of sediments were scraped into plastic vials and fixed with formalin. The remaining sediment in each pot was sieved through a 1.3 mm mesh to retrieve all bivalves and bivalve predators. Bivalves were separated from the preserved sediment using a flotation technique utilizing dense sucrose solu- tions. Peak settlement of Mva occurred from mid to late July in south- western Maine and from late August to early September in eastern Maine. Average abundances were generally two orders of magni- tude greater in southwestern Maine sites than in those in eastern Maine. Preliminary evidence provides no strong indication that predator e.xclusion or tidal height had consistent effects on initial National Shellfishcries Association, Baltimore, Maryland Abstract . 1996 Annual Meeting, April 14-18. 1996 509 settlement and immediate (less than two week) Mya survival. Ap- parently the ditlcrence in abundance between southwestern and eastern Maine Mya populations begins prior to lar\al settlement from the plankton. ESTIMATES OF RECRUITMENT AND ADULT ABUN- DANCE IN THREE FLORIDA POPULATIONS OF BAY SCALLOPS (ARGOPECTE\ IRRADIAXS). Dan C. Marelli,* William S. .Arnold, Catherine Bray, and Melissa Harrison, Florida Department of Environmental Protection, Florida Marine Research Institute. 100 8th Avenue SE. St. Petersburg, FL 33701 . Bay scallops occur in a series of disjunct populations along the Florida Gulf coast from Pensacola to Florida Bay in the Florida Keys. Each population is associated with a coastal basin, and these basins undoubtedly intluence the larval-retention mecha- nisms responsible for the interannual maintenance of bay scal- lop populations. Recruitment in three Florida populations of bay scallops (Argopecten irradians) was examined using artifi- cial spat collectors during 1994 and 1995. We have conducted surveys of adult densities and distributions in the same popula- tions to examine the relationship between recruitment and adult population characteristics within and between basins. In 1994. we examined gonadal development to determine the relationship between spawning and recruitment in each of the populations. We have not seen. a strong relationship between adult densities and levels of recruitment to artificial collectors in any of the three populations examined. Within basins, a stronger relation- ship appears to exist between recruitment and strength of the subsequent adult year-class. These data suggest that recruitment monitoring may allow us to make short-term predictions of adult bay scallop stock levels in Florida populations. Further, we speculate that this predictive ability could become important in making management decisions regarding recreational-harvest reg- ulations. parameters throughout the year than the LI oysters. Multiple spawning events at each tidal height were also witnessed. There- fore, it is concluded that both prolonged maturation and multiple spawnings account for the extended spawning season. Sex ratios differed at the two tidal heights. Females were proportionally more abundant among oysters in high intertidal areas (3.45:1-HC, 3.12: 1-SK vs 1.95:1-HC. 1.85;I-SK at HI and LI, respectively). The higher proportion of LI males was attributed to stresses on the oysters induced by a probable combination of predation, siltation. disease, and competition. These oysters, therefore, would have to allocate resources to repair and maintenance rather than gamete formation. ANALYSES OF GONADAL CYCLING BY OYSTER BROODSTOCK, CRASSOSTREA VIRGINICA (GMELIN), IN LOUISIANA. John E. Supan,* Office of Sea Grant Devel- opment. L.S.U.. Baton Rouge. LA 70803: Charles A. Wilson, Coastal Fisheries Institute. L.S.U., Baton Rouge, LA 70803. Oysters held nearshore in Caminada Bay during the summer, typically exhibit hypertrophic gonads with prominent genital ca- nals beneath translucent mantle tissue about four weeks post- hatchery spawning, indicating recycling. Broodstock (N = 200) were analyzed histologically over a two year period to document such gametogenesis, using Gonad/Body Ratios (GBR) and devel- opmental stages. Ten oysters were randomly selected from a broodstock pool prior to each spawning attempt, and monthly during the winter-spring of 1992. As expected, the mean GBR before successful spawning attempts was significantly greater (P =s 0.05) than the mean GBR before unsuccessful attempts. A dramatic drop in the percent occurrence of the advanced spawning and regression stage from May to June, a steady spawning stage occurrence, and fluctuations in the percent occurrence of early and later developmental stages during the summer months illustrates gonadal recvclins during June-October. VARIATIONS IN GAMETOGENESIS AND SEX RATIOS IN OYSTERS ALONG AN INTERTIDAL GRADIENT. Francis X. O'Beirn* and Randal L. Walker, Shellfish Aqua- culture Laboratory, University of Georgia. Marine Extension Ser- vice. 20 Ocean Science Circle, Savannah. GA 31411-1011. The spawning seasons of oysters in the southeastern U.S. extends from April through September. This study explored the possibility that different gametogenic maturation rates along an intertidal gradient was responsible for this prolonged sea- son. Twenty oysters were taken on a biweekly basis from two ti- dal heights (high intertidal HI and low-intertidal LI) at each of two sites (House Creek HC and Skidaway River SK) in Was- saw Sound, Georgia, from June 1993 to September 1994. Game- togenic condition was evaluated by histological analysis of the gonads and image analysis. There was little or no retardation in gametogenic maturation and spawnings in the HI oysters. Also, the HI oysters tended to maintain higher gametogenic THE EFFECT OF SALINITY CHANGE ON THE SYN- CHRONY OF POLAR BODY DEVELOPMENT IN FERTIL- IZED OYSTER EGGS (CRASSOSTREA VIRGINICA [GME- LIN]). John E. Supan. Office of Sea Grant Development. L.S.U., Baton Rouge, LA 70803; Charles A. Wilson,* Coastal Fisheries Institute, L.S.U., Baton Rouge, LA 70803; Standish K. Allen, Jr., Haskin Shellfish Research Lab.. Rutgers University. Port Norris. NJ 08349. Broodstock, acclimated (1 week) to 13. 20. and 30%. were strip spawned at source salinity and the resultant eggs exposed to 3 different salinities (10, 20, and 30%) to evaluate the effect of rapid salinity change on egg development (synchrony = [the num- ber of embryos observed at metaphase I) — [the number of em- bryos observed past metaphase I] -^ [the total number of embryos counted]). Bonferroni pairwise comparison procedures were used to test for significant differences (a = 0.001) between mean syn- chrony levels of treatment* broodstock interactions at mean de- 510 Abstract. 1996 Annual Meeting, April 14-18. 1996 National Shellfisheries Association, Baltimore, Maryland velopment time. A covariance model (synchrony = treatment sa- linity I broodstock salinity | development time) proved to be ap- propriate for defining the relationship between synchrony and changing salinity. High levels and rates of synchrony were achieved when the treatment salinity ^ broodstock salinity, except when the broodstock salinity was 13%. THE EFFECT OF CYTOCHALASIN B (CB) DOSAGE IN THE SURVIVAL AND PLOIDY OF CRASSOSTREA VIR- GINICA (GMELIN) LARVAE IN LOUISIANA. John E. Su- pan,* Office of Sea Grant Development, L.S.U.. Baton Rouge. LA 70803: Charles A. Wilson, Coastal Fisheries Institute, L.S.U., Baton Rouge, LA 70803; Standish K. Allen, Jr., Haskin Shellfish Research Lab., Rutgers University, Port Norris, NJ, 08349. Survival and ploidy of D-stage oyster larvae were determined following the rearing of embryos exposed to CB dosages of 0.5 mg/l, 0.25 mg/l and 0.125 mg/l for 10-15 mmutes. with 0.05% DMSO and ambient seawater as controls. Since timing did not permit true replication, the experiment was conducted three times on the same day with the same procedures and partially stripping the same male oysters; only different females were used. Treat- ments began when about 50% of the eggs reached PBl (24—31 min). Embryos were reared for 48-hrs at ambient temperature. Mean triploid percentages were 13% ± 6.7% (0.125 mgCB/l), 61.8% ± 6.2% (0.25 mgCB/1), and 68.2% ± 14.1%: (0.5 mgCB/ 1). No significant difference (P =s 0.05) in mean survival was found between the three CB treatments. Significant differences in mean survival between the three experiments (all dosages and controls combined) implies variability due to different sources of eggs. A COMPARISON OF ARTIFICIAL SPAT COLLECTORS IN THE WESTPORT RIVER, MA. Karin A. Tammi* and Michael A. Rice, Department of Fisheries Animal and Veterinary Science, University of Rhode Island, Kingston, Rl 02881; Wayne H. Turner and Bethany A. Starr, Water Works Group, P.O. Box 197 Westport Point, Ma. 02791. Since 1993, the Bay Scallop Restoration Project has utilized 5,200 artificial spat collectors (>4 mm plastic mesh onion bags containing monofilament) for stock enhancement of the bay scal- lop, Argopecten irradians in the Westport River, Massachusetts. Researchers observed that poor scallop recruitment estimates may be attributed to crab predation and fouling inside the collectors. Research conducted in 1995 compared the performance of the original onion bag collector with that of a commercial fine (1.5 mm to 3.0 mm) mesh collector containing a poly-ethylene tube (40 cm X 80 cm) as the settlement substrate. Longlines consisting of 20 collectors, 10 of each bag type, were deployed at Coreys Island for a period of 4 weeks. After 1 month soaking time, longlines were harvested to assess scallop recruitment, crab abundance and fouling in each bag. The fine mesh collector displayed a signifi- cantly (p < 0.05) higher recruitment with a total of 887 scallops compared to 278 from the onion bags (50 collectors each). The fine mesh collector displayed a higher recruitment estimate aver- aging 18.1 scallops per collector, compared to 5.9 scallops per onion bag. Fine mesh bags had fewer mud crabs, Panopeus spp. < 1 mm in carapace length compared to the onion bag which had more crabs of larger size. Fouling was similar for both bag types, but the fine mesh collector appeared to have more siltation and algal fouling than the onion bag. This study indicates that the onion bag made from donated materials is a poorer spat collector design because it allows mud crab colonization, thus increasing the predation on newly settled scallops and the monofilament may settle when suspended in the water column, reducing the surface area available for larval settlement. Scallop recruitment estimates based on onion bag collectors may underestimate actual recruit- ment rates of scallops. HISTOLOGICAL STUDY OF REPRODUCTION IN AR- GOPECTEN VENTRICOSUS. Janzel R. Villalaz,* Departa- mento de Biologia, Acuatica, Universidad de Panama, Panama. A laboratory study was carried out in Delaware to observe changes in reproduction of Argopecten ventricosus by using his- tological techniques in gonads. During 66 days, combinations of monocultures (50:50) of C-ISO and CH-1 were added daily to a tank with filtered and aerated seawater. Salinity and temperature of the water were measured with a salinity-temperature probe meter (YSl). Phytoplankton densities were recorded by direct count with a hematocytometer. This study is a contribution to the reproductive biology of A . ventricosus and fisheries management of the tropical scallop. MARINE GENETICS NUCLEAR DNA MARKERS FOR CRASSOSTREA SPECIES IDENTIFICATION. Patrick M. Gaffney,* College of Marine Studies, University of Delaware, Lewes, DE 11958; Francis X. O'Beirn, Department of Fisheries and Wildlife, Virginia Poly- technic Institute and State University, Blacksburg, VA 24061- 0321. Unambiguous diagnostic methods for identification of cupped oyster {Crassostreu) species at all life stages are useful in a variety of purposes. These include genetic improvement programs involv- ing hybridization or gene transfer, conservation of endangered broodstocks, and ecological monitoring of exotic species inva- sions. We present methods for identification of five Crassostrea species (C. virginka. C. gigas. C. oriakensis {riviilaris), C. sika- mea and C. rhizophorae) by restriction fragment length polymor- phism (RFLP) analysis of both mitochondrial (I6S rDNA, cy- tochrome oxidase 1) and nuclear ribosomal genes (28S, lTS-1 and ITS-2) amplified by the polymerase chain reaction (PCR). We describe methods for analysis of individual eggs and larvae as well National Shellfisheries Association, Baltimore, Maryland Ahstnia. 19% Annual Meeting. April 14-18, 1996 511 as adult tissues, including hemolyniph, which can be used for nondestructive identification of oysters. Wc illustrate the use of these methods to identify viable interspecific hybrids (C. gigas x C. silkamea and C. gigas x C. ariakensis) and to document un- successful attempts at hybridization of C. virginica with C. gigas and C. ariakensis. GENETICS OF SEX DETERMINATION IN CRASSOSTREA OYSTERS: A SINGLE LOCUS MODEL. Ximing Guo* and Standish K. Allen, Jr., Haskin Shellfish Research Laboratory, Institute of Marine of Coastal Sciences, Rutgers University, B-8, Port Norris, NJ 08349; Dennis Hedgecock, Bodega Marine Lab- oratory, University of California at Davis, Bodega Bay, CA 94923; William K. Hershberger, School of Fisheries, University of Washington, Seattle, WA 98105; Kenneth Cooper, DBI Con- sulting, 24888 Taree Drive NE, Kingston, WA 98346. A unique feature of sex in the Crassostrea oysters at the co- existence of dioeciousism, sex change and functional hermaphro- ditism. To determine whether such a system is genetically con- trolled, we analyzed the sex ratio of over 100 factorial and nested families of the Pacific oyster. Crassostrea gigas Thunberg. The overall female percentages of one. two and three-year old oysters were 37%. 55% and 75%. respectively. The increasmg female percentages with time suggest that there were a significant pro- portion of oysters that matured as males first and changed to fe- males in later years. Detailed analysis of family sex ratio revealed significant family differences and parental effects, suggesting sig- nificant genetic control. Further, family sex ratios tended to dis- tribute in four fragmented groups. Those and other data from the literature could be explained by a single locus model of sex de- termination. In this single locus model, there are two alleles: allele Y for maleness and X for femaleness. so that YY is true male. XX is true female, and XY matures as male first and can change sex later on. The three genotypes can produce four types of families under random mating. The sex change of XY oysters may be further influenced by other genetic and environmental factors, which often makes the four types of families less distinctive. In- terestingly, this model of sex determination can account for the evolution of both XY and ZW types of sex determination, depend- ing which allele gains dominance. HYBRID VIGOR IS PERVASIVE IN CROSSES AMONG INBRED LINES OF PACIFIC OYSTERS. Dennis Hedge- cock.* University of California, Davis, Bodega Marine Labora- tory, Bodega Bay, CA 94923-0247. The genetic and physiological bases of hybrid vigor (heterosis) are poorly understood even for major crops. Two alternative ge- netic explanations for heterosis have co-existed for nearly 80 years: dominance and overdominance. In the 1980s, this debate was renewed over numerous reports for bivalve molluscs that in- dividual heterozygosity at allozyme-coding loci was positively correlated with fitness-related traits, primarily growth rate. Be- cause this debate is not likely to be resolved without experiments, we have taken a classical approach to the study of heterosis in the Pacific oyster Crassostrea gigas. controlled crosses among inbred lines. In such mating experiments, heterosis (or potence, h ) for one or more traits can be defined and quantified as Q/L > 1 .0. where L is the difference between the trail values of the two parental inbred lines and Q is twice the deviation of the hybrid from the mid-parent value (Griffing 1990 Genetics 126:753). Over 50 inbred lines of Pacific oyster have now been initiated by selfing of simultaneous hermaphrodites, self-fertilization with cryopreserved sperm after sex-reversal, or by brother-sister mat- ings. Two sets of 2 x 2 crosses were made in both 1993 and 1994 and two sets of 3 x 3 crosses were made in 1995. Observations to date of larval survival and growth (or size-at-age) in larval, juve- nile, and adult stages indicate that non-additive gene action and heterosis for these traits are pervasive. Potence values for growth or size-at-age are always significantly greater than zero and rarely less than 1 .0; in the majority of cases h^ is significantly greater than 1.0, indicating heterosis. These observations alone do not discriminate between the dominance or overdominance hypothe- ses, although one case of negative heterosis suggests a third hy- pothesis, epistasis. The F, hybrids from these crosses are being reared in Tomales Bay, CA. Mapping of quantitative trait loci (QTL) for heterosis will be done by following the segregation of protein and DNA markers and their associations with growth in the F, and backcross gener- ations, as described in the abstract of McGOLDRlCK and HEDGECOCK. Whether these QTL have dominant, overdomi- nant or epistatic effects on growth can also be assessed, so this study should permit resolution of the causes of allozyme heterozy- gosity-fitness correlations. Ultimately, we want to know whether or not non-additive gene action will retard or prevent response to family selection and implicate crossbreeding as an important com- ponent of a genetic improvement program for farmed Pacific oys- ters. MITOCHONDRIAL DNA VARIATION WITHIN AND AMONG LARVAL COHORTS OF PACIFIC OYSTER, CRASSOSTREA GIGAS, DETECTED BY PCR-SSCP ANAL- YSIS. Gang Li* and Dennis Hedgecock, Bodega Marine Labo- ratory. University of California at Davis. P.O. Box 247, Bodega Bay, CA 94923. Detailed studies of genetic composition within and among co- horts of larvae produced by a semi-isolated population of Pacific oysters in Dabob Bay, Wa. are needed to test the hypothesis that large variance in reproductive success causes a previously reported 10''-fold discrepancy between effective and actual population sizes. We cloned and sequenced part of the mitochondrial genome of Crasso.strea gigas and developed PCR primers to amplify, from individual larvae, four fragments totalling about 2.3 kb in length, or 13% of the genome. Each fragment was digested into 512 Abstract, 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore. Maryland smaller pieces and screened for nucleotide-sequence variation by methods for detecting single-strand conformation polymorphism (SSCP). Three temporal plankton samples from Quilcene Bay and two from north Dabob Bay (total N = 519) were surveyed by PCR- SSCP. A common haplotype is shared by 70-84% of all larvae sampled, so Monte-Carlo contingency chi-square methods are used to test the independence of haplotype frequency and sample. Quilcene Bay larval samples are homogeneous, but the north Da- bob Bay larval samples are not. Haplotype frequencies in a cohort of larvae produced after mid-August in north Dabob Bay differ from those in a cohort that appeared in early August in both sites. ATTEMPTED HYBRIDIZATION OF EASTERN AND PA- CIFIC OYSTERS USING BRIDGING CROSSES. Suifen Lyu,* Standish K. Allen, Jr., and Gregory A. Debrosse, Haskin Shellfish Research Laboratory. Rutgers University. Port Norris. NJ. 08349; Patrick M. Gaffney, College of Marme Stud- ies. University of Delaware, Lewes. DE 19958. Past attempts to cross the eastern oyster {Crassostrea virginica) with either of the Pacific oysters, Crassostrea gigas or C. rivularis have failed. In plants, where hybrids are not obtainable by direct means, hybridization can sometimes be carried out using new races or indirectly through bridging crosses; crosses within or be- tween species meant to bridge the incongruity of species. The objective of this research was to test the use of bridging crosses in breaking the incompatibility between eastern and Pacific oysters. Initial experiments were done on the effect of sperm density on fertilization to determine appropriate sperm densities for hybrid crosses. About twice as much heterologous sperm is needed. For attempts with bridging crosses, 17 different varieties of oysters from three species were used. Moderate to high fertilization rates were common in most hybrid crosses. Straight-hinge larvae were obtained from all, but none survived in any hybrid combination. The effects of varying degrees of heterozygosity among the bridg- ing crosses were tested for their effect on fertilization and survival. Highly heterozygous hybrid crosses were no more effective in breaking down hybrid barriers than simple crosses. Apparently the genetic distance, and consequently the incongruity, between east- em and Pacific oysters is too great. MICROSATELLITE MARKER DEVELOPMENT IN THE PACIFIC OYSTER (CRASSOSTREA GIGAS): VARIABIL- ITY TRANSMISSION, LINKAGE AND QTL MAPPING. Daniel J. McGoldrick* and Dennis Hedgecock, University of California, Davis-Bodega Marine Laboratory. P.O. Box 247 Bodega Bay. CA 94923. The development of classes of highly variable, nuclear markers that can be assayed by the polymerase chain reaction (PCR) com- pliments other areas of research including physiological measure- .Ticnt. population genetics, the inheritance of quantitative traits, and -.s integral to addressing a wide range of basic and applied biological questions e.g. "What is genetic basis of hybrid vigor in Pacific oysters?'". To expand the usefulness of molecular ap- proaches in addressing this question, I have cloned and sequenced several microsatellite inserts from a size-fractionated, oyster ge- nomic DNA library. Of some 110 positive clones that I have identified, 50% have yielded PCR primer sets. Testing has revealed that these microsatellites are indeed highly variable, derived from single loci, are co-dominant, and simply inherited, making them very suitable for linkage and mapping studies in C. gigas. I report the known linkage relationships among these microsatellites as well as 25 allozyme loci. Lastly, utilizing the best information from 24 backcross and intercross populations. I present developments in generating the quantitative trait locus (QTL) map for hybrid vigor in the Pacific oyster. THE ROLE OF PHYLOGENETIC DISTANCE ON THE DISRUPTION OF DOUBLY UNIPARENTAL mtDNA IN- HERITANCE IN HYBRID MUSSEL [MYTILUS) POPULA- TIONS. P. D. Rawson and T. J. Hilbish, Dept. Biological Sci- ences, University of South Carolina, Columbia, SC 29208. Blue mussels in the Mytilus edulis species complex have a doubly uniparental mode of mtDNA inheritance with separate ma- ternal and paternal mtDNA lineages. Female mussels inherit their mtDNA solely from their mother while males inherit mtDNA from both parents. In the male gonad the paternal mtDNA is preferen- tially replicated so that only paternal mtDNA is transmitted from fathers to sons. Hybridization is common among differentiated blue mussel taxa; whenever it involves M. trossulus doubly uni- parental mtDNA inheritance is disrupted. We have found high frequencies of males without and females with paternal mtDNA among hybrid mussels produced by interspecific matings between M. galloprovincialis and M. trossulus. In contrast, hybridization between M. galloprovincialis and M. edulis does not affect doubly uniparental inheritance, indicating a difference in the degree to which the mechanisms regulating mtDNA inheritance have di- verged among the three blue mussel taxa. Our data indicate a high frequency of disrupted mtDNA transmission in F, hybrids and also suggest that two separate mechanisms, one regulating the trans- mission of paternal mtDNA to males and another inhibiting the establishment of paternal mtDNA in females, act to regulate dou- bly uniparental inheritance. We propose a model for the regulation of doubly uniparental inheritance which is consistent with these observations. PERFORMANCE OF TRIPLOID OYSTERS, CRASSOS- TREA VIRGINICA, GROWN BY PROJECT O.C. E.A.N. PARTICIPANTS. John Scarpa,* Leslie Sturmer, Everette Quesenberry, Ross Longley, and David Vaughan, Harbor Branch Oceanographic Inst., Ft. Pierce, FL 34946. The use of triploid oysters to offset the poor quality of diploid oysters in the summer was tested in Cedar Key. Florida. One inch triploid oyster seed was distributed to 25 Project O.C. E.A.N. National Shellfishcries Association, Baltimore. Maryland Abstract. 19% Annual Meeting. April 14-18. 1996 513 participants in June 1994. Every three months 5-6 groups were sampled (n = 30 oysters/sample) and compared to either hatchery produced diploids or wild diploids. The triploid group was found to contain >90% triploids. therefore data of diploids was not removed from each data set. A condition index (CI) was calculated using a ratio of dry meat weight to dry shell weight multiplied by 100. In December, after six months of culture, the triploids out- performed the diploids in length (triploid/diploid; 78/68 mm), whole weight (56/40 g). dry meat weight ( 1 .8/1.2 g) and dry shell weight (36.5/24.1 g), but not in CI (5.0/5.0) and prevalence of Dermo (28/7%). During the April sampling it was noted that mor- tality reached virtually 100% for one triploid group and the hatch- ery diploid group, therefore the June sample was compared to wild diploids of similar size and weight. In June the triploids had a higher dry meat weight (4.2/2.1 g) and CI (5.4/3.2) and lower prevalence of DERMO (65/96%). Although variability was evi- dent among triploid groups, indicating local environmental differ- ences as well as differences in culture practices between growers, the value of triploidy for producing highly quality oysters in the summer in Florida is evident. graphic range is unknown. In this study scallops from Massachu- setts, North Carolina, Florida and Texas were collected, spawned and the offspring reared in a common environment to determine if scallops raised under similar conditions exhibited the morphology expected given the geographic origin of their parents. Significant differences among populations were indicated by ANOVA in both wild-caught (13 of 14 morphological characters) and cultured ( 1 1 of 14 characters) scallops. Principal components analysis resulted in the clustering of individuals according to geographical origin. even when scallops were reared in a common environment. The morphological characters most influential in the clustering were plical width, plical number, interplical distance, and valve con- vexity. Geographical variation in morphology appears to have a strong genetic basis and reflects significant genetic differentiation among geographically separated populations of bay scallops. OYSTER DISEASE RESEARCH PROGRAM DGGE REVEALS ADDITIONAL POPULATION STRUC- TURE IN AMERICAN OYSTER {CRASSOSTREA VIRGIN- ICA) POPULATIONS. Jeffrey R. Wakefield* and Patrick M. Gaffney, College of Marine Studies. University of Delaware, Lewes, DE 19958. Sequence variation in the mitochondrial large subunit (16S) ribosomal gene of the American oyster [Crassostrea virginica) was investigated utilizing denaturing gradient gel electrophoresis (DGGE) and direct sequencing methods. The complete sequence of a 360 base pair fragment was characterized in 205 individuals from 21 populations. Three haplotypes (Gulf Coast, South Atlan- tic, and North Atlantic) accounted for 97% of oysters sampled from Maine to Mexico and displayed a high degree of geographic structuring. In contrast, a sample from Prince Edward Island (East River) Canada exhibited seven haplotypes with no single haplo- type found in more than 5 of the 25 individuals assayed. We present a phylogeographic interpretation of these findings and dis- cuss their implications for aquacultural operations. GEOGRAPHIC VARIATION IN MORPHOLOGY OF THE BAY SCALLOP, ARGOPECTEN IRRADIANS (LAMARCK). Ami E. Wilbur* and Patrick M. Gaffney, College of Marine Studies, University of Delaware. Lewes, DE 19958. The bay scallop, Argopecten irradians. exhibits extensive vari- ation in morphology among geographically separated populations, resulting in the recognition of three major subspecies (A. i. irra- dians, A. i. concentricus. A. i. umplicostatus). The extent to which the morphological variation results from differing environ- mental conditions throughout the bay scallop's considerable geo- GLYCOSIDASES IN PERKINSUS MARINUS: PURIFICA- TION AND CHARACTERIZATION OF (J-D-GLU- COSIDASE. Hafiz Ahmed* and Gerardo R. Vasta, Center of Marine Biotechnology, University of Maryland Biotechnology In- stitute, 701 E. Pratt St.. Columbus Center, Suite 236, Baltimore, MD 21202. The mass mortalities of oysters (Crassostrea virginica) in the Chesapeake Bay are due to infection by the apicomplexan parasite Perkinsus marinus. It remains unclear how the parasite gains entry into the host and avoids destruction by phagocytic cells, as well as the pathogenesis mechanisms. In prokaryotic and eukaryotic microorganisms, glycosidase activity has been associ- ated with enhanced virulence. We have detected several cell sur- face glycosidases in in vitro propagated P. marinus, that may play a role in the host-parasite interactions. Of sixteen glycosidase ac- tivities tested in P. marinus cell extracts, activity was found for (in decreasing order) (i-glucoside, p-xylosidase. N-acetyl (J-glu- cosaminidase and N-acetyl p-galactosaminidase. whereas in the spend medium, the maximum activity was observed for N-acetyl P-glucosaminidase. The enzymatic activity was optimal at the fol- lowing pH values: 4.0 for p-glucosidase. 5.0-7.0 for N-acetyl P-glucosaminidase. 4.0-7.5 for N-acetyl P-galactosaminidase. The temperature optimum was 70°C for all. The P-glucosidase was purified to homogeneity by anion-exchange chromatography followed by cation exchange chromatography using HQ/H. HS/M and SP/H perfusion columns. The purified enzyme exhibited a native mol. wt. of 66 kDa on HPLC and subunit mol. wt. of 70 kDa on SDS-PAGE suggesting its monomeric nature. The K^ and V,„^, values with p-nitrophenyl pD-glueoside as the substrate at pH 5 and 37°C were 0.45 mM and 20.8 umol. min respectively. mg 514 Abstract. 1996 Annual Meeting, April 14-18, 1996 National Shellfisheries Association. Baltimore, Maryland HETEROPLOID MOSAICS AND REVERSION AMONG TRIPLOID OYSTERS. CRASSOSTREA GIGAS. FACT OR ARTIFACT. Standish K. Allen. Jr.* and Ximing Guo, Haskin Shellfish Research Laboratory, Rutgers University, Port Norris, NJ, 08349; Gene Burreson and Roger Mann, Virginia Institute of Marine Science. Gloucester Point. VA 23062. Triploids have been proposed for population control for intro- duction and testing of non-native species and for release of genet- ically modified organisms. This type of population control is es- pecially important for marine systems where containment is nearly impossible. In the first field trial of its kind, certified triploid Crassostrea gigas were tested for disease resistance in Delaware and Chesapeake Bays. Certified triploids were obtained by biop- sying about 1500 putative triploids, rejecting diploids, heteroploid mosaics, and those that were ambiguous. After a season of disease challenge in the field, about 15 and 20% of supposed triploids were, in fact, heteroploid mosaics as determined by flow cytom- etry. Mosaicism was expressed among all tissues within an indi- vidual, generally. Flow cytometric results were cross checked with karyology. This and other evidence suggests that heteroploid mo- saics may have arisen as a consequence of reversion of triploids to a mosaic state. A tentative model for reversion of triploids invokes tripolar spindle formation amidst mitoses of triploids cells fol- lowed by differential cell division of the stem (diploid) cell relative to the triploids. Reversion may be a function of a single cell population, namely hemocytes. GROWTH AND TIMING OF JUVENILE OYSTER DIS- EASE (JOD)-INDIICED MORTALITY OF CRASSOSTREA VIRGINICA IN THE DAMARISCOTTA RIVER. ME. USA. Ryan B. Carnegie* and Bruce J. Barber. Department of Ani- mal, Veterinary, & Aquatic Sciences. University of Maine, Orono. ME 04469: Christopher V. Davis, Darling Marine Center and Department of Animal, Veterinary, & Aquatic Sciences, Uni- versity of Maine, Walpole. ME 04573. This study provides insight into growth and the timing and extent of JOD-induced mortality of juvenile Crassostrea virginica under conditions typical of commercial culture in Maine. Repli- cate groups of 500-1000 oysters 2-3 mm in size were deployed biweekly from May 23 through August 31. 1995. in floating trays at a commercial least site on the Damariscotta River. Growth and mortality were monitored weekly through September, and bi- weekly thereafter. Characteristic symptoms of JOD, a cupping of the left valve and chonchiolin deposits on inner valve surfaces, were observed in the final week of July; heavy mortality occurred three weeks later. Survival was highest (<20% cumulative mor- tality (CM)) in the group deployed on May 23, and survival was lowest in the groups deployed on July 20 and August 3 (>90'7f CM). Reduced mortality (<10'7r CM) has been seen in the group deployed on August 31. Oyster culturists in the Damariscotta River area will achieve acceptable survival of juvenile oysters by deploying them in May; more information on the survival of the late-August group over the winter is needed before late-summer deployment can be concluded to be a viable strategy for avoiding JOD-induced mortality. INTRACELLULAR AND EXTRACELLULAR LYSOSO- MAL ENZYME ACTIVITIES IN EASTERN OYSTERS (CRASSOSTREA VIRGINICA). Fu-Lin E. Chu,* Aswani K. Volety, and Georgeta Constantin, Virginia Institute of Marine Science, School of Marine Science, College of William & Mary, Gloucester Point, VA 23062. It is generally believed that lysosomal enzymes play a role in host defense. The hemocyte, plasma and tissue lysosomal en- zymes in oysters maintained at six temperatures (T) and salinity (S) conditions: 3 ppt at 10 and 25°C. 10 ppt at 10 and 25°C and 20 ppt at 10 and 25°C were assayed. The plasma and hemocyte lysosomal enzyme activities in oysters collected from high and low S habitats were also compared. Hemocyte lysozyme (L) concentration, and L-aminopeptidase (L-AP), and acid phos- phatase (AP) activities were affected by both T and S. Hemocyte L and LAP were higher at low T and S than at high T and S. However, highest AP activities were noted in oysters at 25°C and 10 ppt. The highest plasma L and L-AP were in oysters at 3 ppt. T did not affect the plasma L concentration, but L-AP activity was higher at 25'^C than at 10°C. AP was not detected in the plasma. Neither T nor S affected the lysosomal enzymes levels or activities in oyster tissues. In both summer and winter months, the plasma L was significantly higher in oysters from low than from high S habitats. In a summer month, while AP was similar be- tween these low and high S habitats, the L-AP was higher in low than high S habitats. In contrast, there was no difference in hemo- cyte L between low and high S habitats in either summer or winter months and the L-AP and AP were significantly higher in high S than low S habitats. Our results also reveal that plasma L tended to be lower in PerkinsKS marinus infected oysters than in unin- fected oysters. COOPERATIVE REGIONAL OYSTER SELECTIVE BREEDING (CROSBREED) PROJECT. Gregory A. De- brosse* and Standish K. Allen, Institute of Marine and Coastal Sciences, Haskin Shellfish Research Lab, Rutgers the State Uni- versity of New Jersey, Port Norris, NJ 08349. Since 1962 the Haskin Shellfish Research Laboratory has been selectively breeding eastern oysters (Crassostrea virgi- nica) for resistance to the parasite (Haplosporidiiim nelsoni) that causes MSX disease. MSX disease resistance was obtained rel- atively rapidly and pedigreed lines are still maintained. Since 1992, synthetic lines, developed from controlled matings of these same pedigreed lines, were begun in response to Dermo disease pressure. The synthetic lines have undergone 1 '/: generations of selection for Dermo resistance. Using the MSX resistance syn- thetic lines as a foundation and with collaboration of four mid Atlantic institutions (Haskin Shellfish Lab, Rutgers University: College of Marine Studies, University of Delaware; Horn Point National Shellfisheries Association. Baltimore. Maryland Abstract. 1996 Annual Meeting. April 14-18. 1996 515 Environmental Laboratory, University of Maryland; Virginia In- stitute of Marine Science. College of William and Mary), the objective of this project is to institute a regional selective breeding program for developing oyster stocks resistant to both MSX and Dermo disease. Experimental groups were deployed to the partic- ipating institutions in August 1995. The experimental design, hatchery production, and progress of the project to date will be discussed. RESISTANCE STUDIES FOR JUVENILE OYSTER DIS- EASE (JOD). Austin C. Farley* and Earl J. Lewis, National Marine Fisheries Service. Oxford. MD 21654; David Relyea. Joseph Zahtila. Frank M. Flower Co.. Oyster Bay. NY 11771; Gregg Rivara, Cornell University. Cooperative Extension. Southold, NY 11971. In a previous study. F, progeny from oyster brood stocks se- lected on the basis of survival of exposure to juvenile oyster dis- ease (JOD) were exposed to the disease at two infective sites in 1994 along with seed oysters (FCT) from naive Connecticut brood stocks. Significant differences in survival warranted further stud- ies. Susceptible progeny. F, resistant progeny, and F, progeny from 1993 brood stocks from the F, generation were produced from June 1995 spawnings at the Frank M. Flower hatchery in Bayville, NY. Seed were placed in nursery raft trays on Aug. 4. 1995 and deployed at 7 sites in the Long Island area on Aug. 28. Evaluations for JOD (size, shell checks, conchiolin prevalence in live and dead oysters, and mortality) were made from weekly or biweekly samples between Aug. 28 and Nov. 6, 1995. No mor- tality was seen on Aug. 28 at the Flower site. Size culled FCT susceptible runts ( 16-20 mm) had mortalities of 13% on Sept. 12, 67% on Sept. 28, 74% on Oct. 17. and 75% on Oct. 30, 1995. Size culled resistant F, and F, runts had mortalities of 0 to 3% during this time. Unculled oysters had a maximum mortality of 15% in the FCT seed and <2% in the F, and F, seed over the same time period. At another Long Island Sound site, unculled FCT seed had mortalities of 30% on Sept. 18, 52% on Oct. 3, 62% on Oct. 16, and 72% on Oct. 30, 1995. The F, and F, seed had mortalities of 3%, 2%, 2%, 13%, and 3%, 4%., 3%, 15%, respec- tively, on the same dates. At 5 Peconic Bay sites, mortalities of 43-72% were seen by Oct. 30 in the FCT seed, while the F, and Fi seed mortality was between 1% and 15%. Conchiolin prevalences averaged 407f in FCT live seed, and 5% in the F, and F, live seed. The results of this study demonstrate that seed oysters from JOD surviving brood stocks are 7 to 25 times better able to survive exposure to this disease than susceptible control populations. Use of these brood stocks on a commercial scale has brought produc- tion at the Frank M. Flower Co. back to pre-JOD levels for oys- ters. INHIBITION OF PERKINSUS MARINVS IN VITRO PRO- LIFERATION BY HETEROLOGOUS PLASMA. Julie D. Gauthier* and Gerardo R. Vasta, Center of Marine Biotechnol- ogy, University of Maryland Biotechnology Institute, Columbus Center. Suite 236. 701 E. Pratt Street. Baltimore. MD 21202. The endoparasitic protozoan Perkinsus marinus (Phylum Api- complexa; Class Perkinsea) is considered the primary cause of mass mortalities of the eastern oyster Crassosstrea virginica, and no natural resistance to the disease has been described. The in vitro culture of the pathogenic stage of P. marinus has provided a unique opportunity to examine its susceptibility to recognition and effector defense mechanisms operative m refractory bivalve spe- cies. We present the effect of medium supplementation with plasma from ( 1 ) uninfected to heavily infected oysters (C. virgin- ica). (2) uninfected disease resistant west coast oysters (C. gigas and C. riviilaris) and (3) other east coast bivalves (Mytilus edulis, Mercenaria mercenaria and Anadara oralis) that are naturally exposed to the pathogen but show no signs of disease. Our results demonstrate a significant (P < 0.05. Fisher PLSD) decrease in growth of P. marinus in the presence of plasma from infected vs. uninfected C. virginica. Plasma (20%) from heavily infected oys- ters inhibited growth by 32% relative to control with no plasma supplementation. Plasma from other bivalves inhibited growth in a dose-dependent manner. Growth was significantly reduced (P < 0.05. Fisher PLSD) in media supplemented with M. edulus (5%), A. oralis (10%) and M. mercenaria (20%) plasma. The highest inhibitory activity was found in M. edulis: only 5% plasma was needed to reduce growth by 35% compared with the control. Plasma from west coast oysters was not inhibitory; in fact growth was significantly enhanced (P < 0.05. Fisher PLSD) in media supplemented with C. rirularis (35%) or C. gigas (20%) plasma. GENE TRANSFER THROUGH HYBRID PARTIAL GYNO- GENESIS BETWEEN THE PACIFIC AND AMERICAN OYSTERS. Ximing Guo,* Standish K. Allen, Jr., and Patrick M. Gaffney, Haskin Shellfish Research Laboratory. Rutgers Uni- versity. B-8. Port Norris, NJ 08349; College of Marine Studies, University of Delaware, Lewes. DE 19958. The transfer of disease resistance genes from the Pacific oyster to the American oyster is prohibited by post-gametic barriers to hybridization. It is possible that those barriers are caused by only a small number of incompatible genes and can be potentially by- passed with a partial genomic transfer via hybrid partial gynogen- esis. Gynogenesis refers to the development of eggs promoted by genetically inactivated sperm. In partial gynogenesis. the sperm genome is partially inactivated, so that a fraction is incorporated in the gynogenetic development. In this study, we tested the feasi- bility of hybrid partial gynogenesis between the Pacific and Amer- ican oysters. The goal is to use hybrid partial gynogenesis for the transfer of disease resistance between the two species. Hybrid partial gynogenesis was induced by fertilizing Ameri- can oyster eggs with ultraviolet (UV) irradiated sperm from the 516 Abstract. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheiies Association. Baltimore. Maryland Pacific oyster followed by blocking polar body II. UV irradiation was very effective in damaging sperm chromosomes. Partial de- struction of sperm genome was produced at low UV dosages as evidenced by chromosomal fragments. Diploid American oyster embryos with Pacific oyster chromosomal fragments were pro- duced at high efficiencies. However, chromosomal fragments cre- ated by UV irradiation greatly reduced the viability of oyster em- bryos. Only a small number of survivors were obtained from groups where sperm were irradiated at high dosages. Whether those survivors carry any genes from the Pacific oysters will be determined in future analysis. JUVENILE OYSTER DISEASE— TRANSMISSION AND BACTERIOLOGICAL STUDIES. Earl J. Lewis* and Austin C. Farley, National Marine Fisheries Service, Oxford. MD 21654; Ana Baya, Maryland Department of Agriculture. College Park. MD 20740; Eugene B. Small, University of Maryland. College Park. MD 20742. Juvenile oyster disease (JOD) has plagued the aquaculture in- dustry in New York and New England since the late 1980s. This new disease causes devastating mortalities in oysters, Crassostrea virginica. typically less than 30 mm of size. While the identity of the disease agent is unknown, our studies have shown JOD to be a transmissible, temperature and salinity sensitive, waterborne. infectious disease with an incubation period of 3-7 weeks. Experimental studies demonstrated the disease to be transmis- sible using JOD-infected oysters, or particulates filtered from am- bient water at JOD-infected growing sites. Experimentally in- fected oysters consistently reveal the JOD syndrome as seen in naturally infected oysters. Transmission was evident at salinities of 18 ppt and above, but the disease agent was shown to survive in oysters held for 7 months in <5 ppt salinity water and to cause disease after salinity was raised to 26 ppt. Mortalities in JOD- infected oysters were also reduced by maintaining them in low salinity water. Vibrio spp. have been isolated from JOD-infected oysters and water samples in bacteriological studies, but our data do not sug- gest a bacterial cause for this disease. Intracellular bodies, sug- gestive of a protistan parasite, have been observed consistently in histological sections of mantle from JOD-infected oysters. Protists isolated from ambient water at JOD-infected sites and aquaria used in transmission studies are being investigated as the possible caus- ative agent(s). ISOLATION AND CHARACTERIZATION OF MARKER GENES FOR PERKINSUS MARINUS. Adam Marsh, Anita C. Wright,* and Gerardo R. Vasta. Center of Marine Biotech- nology. University of Maryland Biotechnology Institute. Suite 236. Columbus Center. 701 E. Pratt St.. Baltimore. MD 21202. Perkitisus inariinis. an apicomplexan parasite of the Eastern oyster, Crassostrea virginica. is the causative agent of devastating lethal disease when the host is exposed to certain environmental conditions such as increased salinity and temperature P. marinus has been successfully cultured in host-free medium, greatly facil- itating genetic studies of this organism, which were previously limited to sequence of ribosomal RNA. In this report, we present nucleic acid sequence with similarity to p actin. Messenger RNA was extracted from P. marinus growth in the presence of oyster serum and reverse transcribed to generate cDNA. Conserved se- quence from the 5' and 3' ends of human actin nucleic acid se- quence were used to design primers for PCR amplification of cDNA. Limited sequence of the cloned PCR product permitted elaboration of primers specific for P. marinus which were used for further amplification. PCR product was subcloned into pBluscript and sequenced. While invertebrate and vertebrate species share 94—96% amino acid identity for actin. sequence of P. marinus revealed 83 to 84% identity to host (C. virginica) or mouse actin amino acid sequence, respectively. These findings are consistent with other apicomplex- ans, such as Trypanosoma cruzi. which have similar sequence diversity of 82 to 84% amino acid identity to vertebrate, inverte- brate or P. marinus actin sequences. RECONSTRUCTION OF A NATURAL OYSTER BAR IN THE CHOPTANK RIVER USING HATCHERY PRO- DUCED OYSTER SEED. Donald Meritt,* Horn Point Envi ronmental Laboratory. University of Maryland, Cambridge, MD 21613; Kennedy T. Paynter, Department of Zoology, University of Maryland. College Park. MD 20742; Robert Pfeiffer, Oyster Recovery Partnership. Annapolis, MD 20676. The Maryland Oyster Recovery Action plan calls for the use of disease-free hatchery seed in the reconstruction of oyster bars in specific zones of certain Maryland. These zones — designated A, B. & C — are areas in which specific restrictions are enforced. In Zone A, no oyster harvesting is allowed and no dermo or MSX infected seed may be planted. In Zone B. no infected seed may be planted but harvest can occur. Zone C areas have no particular restrictions at the present time. This plan will allow experimental projects to be performed to test how quickly oyster seed will become infected in these areas and how diseases affect oysters in relatively low salinity areas over a period of years. Reconstruction of a natural oyster bar in the Choptank River was initiated in 1995 with funding from NOAA. The project was started with the deposition of 100.000 bushels of dredged fossil oyster shell by the Maryland Department of Natural Resources on a 10 acre portion of a natural bar in Zone A of the Choptank River. This produced a large, hard platform on which the hatchery-reared seed could be planted. Oyster larvae required for the project were produced at the Horn Point Environmental Lab hatchery in three large batches approximately 2 weeks apart from each other. Set- ting was conducted in two 10,000 liter tanks. Spat were held in these tanks 4 to 10 days after settlement then moved to nursery National Shellfisheries Association. Baltimore. Maryland Abstract. 1996 Annual Meeting. April 1-1-18. 1996 517 sites in the Choptank River. After 4 to 6 weeks at the nurser>' sites where the spat grew to approximately 15 mm. they were planted on the prepared oyster bar. By October. 1995. the spat had grown to an average height of 28 mm. Surveys will continue through 1996 to determine growth, survival and infection rates of the oys- ters. EVALUATING EASTERN OYSTER STOCKS FOR RE- SOURCE REHABILITATION. Kennedy T. Paynter,* Depart ment of Zoology. University of Maryland. College Park. MD 20742; Patrick M. Gaffney, College of Marine Studies. Univer- sity of Delaware. Lewes. DE 19958; Donald Meritt, Horn Point Environmental Laboratory. University of Maryland, Cambndge. MD 21613. American oysters from several Atlantic and Gulf coast loca- tions were used as broodstock to produce hatchery lines, which are now being grown out in low- and high-salinity Chesapeake Bay waters. They are being evaluated for growth rate and resistance to MSX and Dermo. In addition, both the hatchery lines and their respective progenitor broodstocks are bemg examined by denatur- ing gradient gel electrophoresis (DGGE) and direct DNA sequenc- ing of mitochondrial genes, in order to determine genetic relation- ships among geographical populations and to obtain genetic mark- ers for discriminating among them. Genetic groups were deployed in September 1995. in the Chester and Choptank Rivers in Maryland and in Mobjack Bay. VA. Initial sizes of the groups were similar, ranging between 13 and 19 mm average shell height. By November. 1995. average shell heights ranged from 31 to 45 mm. In general the oysters were slightly larger at the Mobjack Bay site although some groups at the Maryland sites were larger than their counterparts at Mobjack Bay. Haplosporidium netsoni was not detected in any groups in Octo- ber, 1995, but I to 10 putative Perkinsus marinus cells were detected in a few individuals in a few groups. Initial genetic anal- ysis indicates that the broodstocks can be divided into Gulf Coast, North Atlantic and South Atlantic groups, with mid-Atlantic wa- ters containing both Atlantic types. Information obtained will be used to I ) determine the extent to which geographical populations are genetically distinct stocks; 2) determine which of these stocks would be best for replenishing native populations in disease-ravaged areas; 3) provide genetic markers useful for determining the geographic origin of oysters, for use in management, enforcement and breeding programs. ASSESSMENT OF GEOGRAPHIC VARIABILITY IN PER- KINSUS MARINUS. Jose Antonio F. Robledo,* Adam G. Marsh, Anita C. Wright, and Gerardo R. Vasta, Center of Marine Biotechnology. University of Maryland Biotechnology In- stitute. 701 E. Pratt St.. Columbus Center. Suite 236. Baltimore. MD 21202. Perkinsus marinus is the major cause of mortality of the eastern oyster. Crassostrea virginica. in the Chesapeake Bay. This situ- ation has resulted in a reduction of the oyster production, and better understanding of this organism at the molecular level is needed. PCR primers derived from a non-transcribed spacer (NTS) domain between the 5S and 17S small subunit of rRNA sequence, previously shown to be specific for P. marinus. were used to compare P. marinus sequence from infected Chesapeake Bay and Gulf Coast oysters. Total DNA ( 1 |J.g) was extracted from oysters (n = 8) obtained from either Maryland or Louisiana. PCR prod- ucts of the amplified 307 bp target region were cloned into pBlueScript. and several clones of each isolate carrying the inserts were sequenced. P. marinus sequence of the NTS domains from all Maryland oysters were identical. The nucleotide sequence of the NTS from Louisiana isolates showed greater variability. Most of the sequence dissimilarity (four positions) was concentrated in a region of nine nucleotides. Two of the Louisiana samples were identical to sequence of P. marinus from Maryland; however, within the nine-base region, the other six sequences exhibited only 55.5'7c nucleotide identity to Maryland sequence but were identical to each other. Differences between NTS domain rRNA sequences may represent genetic diversity within P. marinus populations in various geographic areas and might provide a tool to better under- stand the relationship between P. marinus strains isolated from oysters at different locations. ACID PHOSPHATASE: A VIRULENCE FACTOR OF THE PROTISTAN PARASITE, PERKINSUS MARINUS AGAINST HOST, OYSTER'S DEFENSE? Aswani K. Volety* and Fu-Lin E. Chu, Virginia Institute of Marine Science, School of Marine Science. College of William & Mary. Gloucester Point. VA 23062. Acid phosphatase (AP) in parasites, has been postulated to play a role in evasion of the host defense by dephosphorylation of host phosphoproteins and/or suppression of the oxygen intermediates released by the host phagocytes. The effect of temperature (4. 12. 20 and 28°C) and osmolality (400, 570 and 840 mOsm/kg) on extracellular AP secretion by the oyster parasite, Perkinsus marinus was investigated in vitro. In addition, ultrastructural localization of acid phosphatase activity in P. marinus. was also examined. Temperature significantly affected AP secretion by P. marinus (p < 0.0001). AP activity in P. marinus in- creased with the increase of temperature (p < 0.05). The extra- cellular AP secretion was P. marinus cell density-dependent (p < 0.00 1). Increasing temperatures resulted in increased proliferation of P. marinus cells. Similarly, osmolality signifi- cantly affected extracellular AP secretion by P. marinus (p < 0.0001). AP secretion was higher in P. marinus cells cultured at 400 and 570 mOsm/kg than those cultured at 870 mOsm/ kg media. In the ultrastrueture study, intense AP activity was found in the nucleus of the parasite. Based on the activity and distribution of AP in the nucleus. AP may aid the parasite in avoiding host defense and maybe involved in cell cycle reg- ulation. 518 Abstract. 1996 Annual Meeting, April 14-18, 1996 National Shellfisheries Association. Baltimore, Maryland SHELLFISH NEOPLASIA GONADAL NEOPLASIA IN HARD CLAMS {MERCE- NARIA SPP.) FROM THE INDIAN RIVER LAGOON, FLORIDA. W. S. Arnold* and T. M. Bert, Florida Department of Environmental Protection. Florida Marine Research Institute, 100 Eighth Avenue S.E., St. Petersburg. FL 33701; D. M. Hes- selman, United States Food and Drug Administration. P.O. Box 21077. Charleston, SC 29413; N. J. Blake, Department of Ma- rine Science. University of South Florida. 140 Seventh Avenue S., St. Petersburg. FL 33701. Within the past 20 years, numerous cases of neoplastic disease in bivalve molluscs have been reported. These neoplastic diseases have typically been of hemic origin, but until recently only a very few cases of gonadal neoplasia m bivalve molluscs had been re- ported. Of those cases in which the neoplasm was of gonadal origin, only one occurrence had been documented for the hard clam Mercenaria . However, dunne a recent three-year study, we documented a relatively high incidence of gonadal neoplasia in hard clams from the Indian River lagoon. Florida. The Indian River hard clam population is composed of approx- imately 68% M. mercenaria . 4% M. campechiensis. and 28% hybrid genotypes. The incidence of gonadal neoplasia is substan- tially higher m hybrids than in either of the two pure species. Although there is a locational component to the incidences of gonadal neoplasia, the pattern is attributable to the proportion of hybrids at that location rather than to any locational characteristics per se. In fact, water quality in the lagoon is generally good, and the fact that we have not found any localized concentrations of neoplastic clams indicates that water-borne carcinogens are not present. The frequency of occurrence of gonadal neoplasia was different for males and females; we observed an overall lower frequency of the disease in males than in females, and the disease was most common in males of intermediate age but in females of very old age. However, hybridity rather than environmental or other biological factors appears to determine susceptibility, impli- cating a genetic mechanism in the etiology of the disease. Genetically mediated gonadal neoplasia may provide an endog- enous mechanism that operates in conjunction with exogenously mediated, genotype-specific growth differences to balance selec- tion on hard clam genotype classes in the Indian River lagoon. This balancing selection acts to maintain the Indian River hard clam hybrid zone in spite of a preponderance of M. mercenaria alleles within the population. GONADAL NEOPLASMS IN MY A ARENARIA. WHAT DO WE KNOW? Bruce J. Barber,* Department of Animal. Veter- inary & Aquatic Sciences, University of Maine, Orono, ME 04469. 1) Gonadal neoplasms have only been found in Mya are- nana from Maine. Prevalence of gonadal neoplasms in adult Claras from Washington County. ME. since September 1993 has ranged from 10% to 43%. Intensity of neoplasms varies from a few. small foci of undifferentiated germ cells (Stage 1). to 50- 100% of gonadal follicles being involved (Stage 2). to invasion and metastasis with loss of tissue architecture (Stage 3), indicating that the disease is progressive and lethal. Clams of both sexes are affected, but females are significantly more likely (P =£ 0.025) to have neoplasms than males. 2) Neoplasms negatively impact gameotogenesis. Female clams with neoplasms produce 66% fewer oocytes than healthy clams (P « 0.001) as the result of direct displacement by tumor cells. In addition, female clams with neoplasms have a significantly lower mean oocyte diameter before spawning and a significantly greater mean oocyte diameter after spawning (P ^ 0.001) than healthy clams, as the result of a gen- eral inhibition of normal oogenesis and spawning. 3) Absolute diagnosis of gonadal neoplasms requires histological examina- tion. Clams with advanced neoplasms can also be identified visu- ally by the presence of a mottled gonad; clams without discemable mottling, however, may have Stage 1 neoplasms. Blood from clams with gonadal neoplasms has levels of glucose and total protein similar to levels in healthy clams. GONADAL NEOPLASIA IN NORTHERN AND SOUTH- ERN QUAHOGS AND THEIR HYBRIDS IN SOUTH CARO- LINA. Arnold G. Eversole,* Department of Aquaculture, Fish- eries and Wildlife. Clemson University. Clemson. SC 29634- 0362; Peter B. Heffernan, Marine Institute. 80 Harcourt Street, Dublin 2. Ireland. Histopathological examination of northern and southern qua- hogs and their hybrids revealed gonadal neoplasia in 47% of the samples (n = 440). Severity of neoplasia was assigned scores from very low ( 1 ) to a high severity condition (5). Average se- verity value was 1 .89, and the low severity category was the most frequently (44%) encountered condition. Highest severity values were observed May-July and September-October and lowest val- ues were in March and August. Shell lengths (SL) of samples were 19.9-61.4 mm and those with neoplasia were 20.1-51.2 mm. Average SL of quahogs with neoplasia were 5^ than the SL of specimens without neoplasia. The SL of hybrids with more ad- vanced stages of neoplasia were > than SL of hybrids without neoplasia. Sex composition of the samples were 39% male, 40% female and 21% undifferentiated. Neoplasia prevalence (59%) and severity (2.22) were highest in the undifferentiated sex category. Average monthly prevalence and severity of neoplasia was > than in either parental species. Gonadal neoplasia was diagnosed in all the hybrid monthly samples and 58% of the parental species monthly samples. Maximum monthly prevalence was 100% in the hybrids and 75% in the parental species. The frequency of gonadal neoplasia in the hybrids increased between 1988 and 1992; aver- age prevalence was 95-100% and severity was 3.41^.21 in 1992. Invasive stages of neoplastic cells were observed in the hepato- pancreas. Examination of gonads from more recent collections National Shellfisheries Association, Baltimore. Maryland Abstract, 1996 Annual Meeting, Apnl 14-18, 1996 519 will be compared with these samples to determine progression of the condition and degree of invasion of other tissues. AN UPDATE ON SOFTSHELL CLAM (MYA ARENARIA) SARCOMA IN THE CHESAPEAKE BAY AND THE DE- CLINING FISHERY. Shawn M. McLaughlin* and Austin C. Farley, National marine Fisheries Service. Southeast Fisheries Science Center, 904 S. Moms St.. Oxford, MD 21654; Christo- pher C. Judy, Maryland Department of Natural Resources. 580 Taylor Ave.. Annapolis. MD 21401. Presumptive hematopoietic neoplasms (sarcomas) of softshell clams. Mya arenana. were rarely observed in Chesapeake Bay populations prior to the first documented epizootic in 1984. Soft- shell clam sarcoma epizootics were subsequently reported in six Chesapeake Bay sites, including Swan Point in the Chester River, during 1984—1988. Subsequent monitoring of the Swan Point site showed sarcoma prevalences of l%'7c in December of 1990. 0% during summer of 1991. and 20'7f in January of 1992. Soft- shell clam sarcomas at Swan Point have continued at low preva- lences (0-39^) since 1992 with small peaks at 12'/'( each in March 1994 and July 1995. The apparent seasonal and cyclic nature of softshell clam sarcoma epizootics suggests another peak in sar- coma prevalences may be imminent at Swan Point. However, actual sarcoma prevalences may be obscured by a reduced sam- pling frequency due. in part, to decreased commercial clamming activity. Softshell clam harvests in the Maryland portion of the Chesapeake Bay have declined from a high of 680.358 bushels m 1964 to a low of 29.616 bushels in 1992 and only reached 37.386 bushels in 1994. The relationship between softshell clam sarcomas and the declining commercial harvest in the Chesapeake Bay is complicated by environmental and biological factors. High prev- alences and intensities of softshell clam Perkinsus sp. and a chla- mydia-like organism have also been observed in Chesapeake Bay softshell clams. In addition, extreme summer temperatures have been associated with high softshell clam mortalities in the Ches- apeake Bay. THE RATE OCCURRENCE OF NEOPLASIA IN CRUSTA- CEA: MYTH OR SAMPLING ARTIFACT. J. Frank Mo- rado* and Donald V. Lightner, National Oceanic & Atmospheric Administration. National Marine Fisheries Service. Alaska Fish- eries Science Center. 7600 Sand Point Way NE. Seattle. WA 98115; Environmental Research Laboratory. University of Ari- zona. 2601 East Airport Drive. Tucson. AR 85706. The Class Crustacea is both an ecologically and economically important taxon that contains roughly 30000 species. It is a highly diverse and successful taxon possessing easily recognizable adult stages such as crabs, lobsters and shrimp to the highly aberrant Rhizocephala and parasitic isopods and copepods. The majority of the Crustacea are aquatic, but many are terrestrial. Survival is dependent upon the production of thousands to millions of progeny that must pass through several larval stages during which large increments in growth occur before they acquire their adult char- acteristics. Two basic tenets for the induction of cancer were presented by Cotran. Kumar and Robhins {.Pathologic Basis oj Disease, 1989). The first simply stated that risk was associated with survival. The second identified cell replication because it is involved in cancer- ous transformation — "regenerative, hyperplastic and dysplastic proliferations are fertile soil for the ongin of a malignant neo- plasm." Crustaceans possess these two criteria, but why is neo- plasia rare in this class of animals, especially when embryos, larvae and adults in aquatic coastal environments are exposed to the same chemical promoters of neoplasia as fish? To date, only a handful of published reports present realistic accounts of neoplastic disorders in crustaceans. Two may be clas- sified as benign neoplasms while the three most recently described cases may be classified as carcinomas. Descriptions of these pro- liferative anomalies as well as their prevalences and distributions will be presented. Probable reasons for the rarity of neoplastic lesions in crustaceans, regardless of their life history stage will be discussed. EFFECTS OF HEMATOPOIETIC NEOPLASIA ON RE- PRODUCTION AND POPULATION SIZE DISTRIBUTION IN THE SOFT-SHELL CLAM. Mary-Susan Potts,* Biology Dept. Northeastern University, Boston, MA 02115. The soft-shell clam, Mya arenana, is susceptible to Hemato- poietic neoplasia (Hn), a disease in which increasing numbers of atypical cells invade the hemolymph and connective tissue of the clam's circulatory, digestive, reproductive and excretory systems. In this study disease effects on the morphology of the clam's reproductive organs were examined. In addition. Hn prevalence was evaluated with respect to sex and as a function of size in the total population and in a cohort of clams. Methods utilized in- cluded field sampling, routine histology and computer image anal- ysis. Comparisons of mean gonadal follicle size and number of follicles were made between normal and Hn clams. Qualitatively. Hn effects on the gonads of both male and female clams were increasingly apparent as abnormal cells filled the clam's connective tissue. As Hn advanced, the gonadal folli- cles became significantly smaller in size but were not reduced in number. While Hn prevalence was unrelated to the sex of the clam, the data supported a relationship between Hn prevalence and size. Clams 40-70 mm had the highest prevalence of Hn. and a cohort of clams showed increasing Hn prevalence as they grew into this disease-susceptible size range. In general the data suggest that Hn reduces the reproductive capabilities of diseased individuals and may directly alter the size distribution of soft-shell clam populations by removing particular size classes through mor- tality. 520 Abstract. 1996 Annual Meeting. April 14-18, 1996 National Shellfisheries Association, Baltimore, Maryland NEOPLASIA AND OTHER POLLUTION ASSOCIATED LESIONS IN MYA ARENARIA FROM BOSTON HARBOR. Roxanna Smolowitz,* Laboratory for Marine Animal Health, U. Of Pennsylvania, Marine Biological Laboratory, Woods Hole, MA 02543; Dale Leavitt, WHOI, Woods Hole, MA 02543. Northeast coastal populations of Mya arenaria. the soft shell clam, are affected by Hematopoietic Neoplasia. High prevalences of this disease has been related to highly polluted waters by some researchers. Therefore Mya was chosen to investigate as a possible biological monitor of pollution in Boston Harbor. Twenty animals were collected from 5 sites in Boston Harbor and 2 control sites in Cape Cod Bay. The occurrence of Hematopoietic Neoplasia in these samples was determined using a neoplasia cell specific an- tibody in a modified immunocytochemical staining method. In addition, paraffin embedded tissues were examined from each an- imal for other possible pathologies. Multivariant statistical meth- ods were used to determine what pathologies were highly corre- lated with animals collected from polluted sites. Out of 33 possible pathologies, the occurrence of the following in Mya were highly correlated with pollution: gonadal inflammation, inflammation of the mantle, gill inflammation, gill hyperplasiaypapilloma, kidney hyperplasia, brown cell accumulation in the renal epithelium, pro- tozoal infection in the kidney and pericardial gland changes. He- matopoietic neoplasia was not positively corrected with site pol- lution. Surprisingly, 5/20 animals from only one polluted site showed a neoplasia not identified previously in Mya. This neopla- sia appeared to be branchial in origin and had numerous metastatic nodules in other tissues. Whether this neoplasia was caused by pollution, an infectious agent or a combination of causes, is not known. INVESTIGATION OF MOLECULAR MECHANISMS OF TUMORIGENESIS IN BIVALVE GONADAL TUMORS. R. J. Van Beneden and L. R. Rhodes, University of Maine, Orono, ME; D. J. Brown, Duke University, Durham, NC; and G. R. Gardner, EPA, Narragansett. Rl. Epidemiological investigations of germ cell tumors of Maine soft shell clams {Mya arenaria) and hard shell clams {Mercenaria spp.) from Florida demonstrate the prevalence of histogenically similar gonadal cancers as high as 40% and 60%, respectively. Human mortality rates due to ovarian cancer from the same areas are significantly greater than the national average and show a rise over the last two decades of the survey. This correlates with the increasing use of herbicides in these areas and in the appearance of significant numbers of tumors of the analogous reproductive sys- tem in the clams. We have investigated the molecular mechanism of tumor induction in the clams. Results of photoaffinity binding studies using the TCDD photoaffinity analog l'"'^-I]-2-azido-3- iodo-7,8-dibromodibenzo-;?-dioxin to detect two cytosolic proteins (28 and 39 kDa) in Mercenaria mercenaria and one (35 kDa) m M. arenaria which specifically bound this ligand. Their role in dioxin toxicity and their relationship to the vertebrate Ah receptor is under investigation. Differential display PCR analysis has re- vealed the presence of differentially-expressed messages in tumors from M. arenaria which have potential roles in signal transduction pathways. EXPRESSION OF THE TUMOR SUPPRESSOR GENE, P53 IN NORMAL AND LEUKEMIC CLAM BLOOD CELLS IN VIVO AND IN VITRO. Charles W. Walker, Sharon A. Key. Joseph E. Mulkern, Shalini Verma, and Jocelin A. Jacobs, Department of Zoology, University of New Hampshire, Durham. NH 03824. Mya arenaria. the soft shelled clam, can develop a diffuse blood tumor or leukemia. The disease is fatal and may deplete commercial clam populations. The only externally obvious signs of the disease are slow decline and death of affected individuals. Thus, the degree of destruction of local clam beds may not be obvious to local fishermen and has not been carefully documented anywhere in the Gulf of Maine. Until now it has been impossible to culture these rapidly dividing cells for use in molecular or other studies. We have developed a method for mass culture and cryo- preservation that should permit more widespread study of these tumor cells. Using our chemically defined medium (modified from Sible et al., 1991) in spinner culture flasks |8-10°C), leukemic cells harvested directly from the clam heart double in number in 40-50 hours and can easily be subcultured. We can also maintain leukemic cells in biofreezers ( - 196°C) for extended periods of time (S3 months) and can subculture recovered cells. In both cases, we have confirmed their identity with the clam leukemia specific antibody lElO (Miosky et al., 1989) and cytology. We have partially cloned the tumor suppressor gene p53 from normal clam blood and gonadal tissues. In the available sequence, 3 of the 5 vertebrate DNA bindmg domains exist and are highly conserved as are the nuclear translocation and dimerization and tetrameriza- tion domains. Based on structural conservation, the p53 gene product should function similarly in down-regulating cell division in clams and vertebrates. When compared with normal clam blood cells from New Bedford Harbor, Massachusetts, our data from in situ hybridizations and cytochemistry demonstrate that expression of clam p53 is depressed in fully developed leukemia cells. This expression pattern is consistent with that seen in vertebrate leuke- mias. Because of the high degree of similarity between p53 and many of the genes involved in mitogenic signal transduction cas- cades between mammals and clams, our studies may have sub- stantial implications for understanding leukemia in mammals. Also, molecular data that we gather about leukemia in clams should point out mechanisms for the diagnosis, treatment and pre- vention of clam leukemia and should lead to an understanding of how the disease is transmitted and promoted in clam populations at risk for this fatal disease. Supported by Hatch Grant #353 to C. W. Walker; McNair, to S. A. Key, SURF to J. E. Mulkern and UROP to S. Verma. National Shellfisheries Association. Baltimore, Maryland Abstrcicl. 1996 Annual Meeting, April 14-18, 1996 521 STOCK ASSESSMENT USING REAL TIME DATA WITH A PC-BASED GIS FOR SHELLFISH MANAGEMENT. James G. Boyd,* Duke Uni versity School of the Environment, Beaufort, NC 28516; William D. Anderson and Guy M. Yianopoulos, South Carolina Depart- ment of Natural Resources, Charleston, SC 29422. Estuarine environments are inherently dynamic systems. Con- sequently, management of estuarine resources is facilitated by tools that allow for rapid evaluation and decision making. Using a pc-based geographic information system (GIS/ArcView2) to dis- play shellfish resources in a portion of the James Island USGS quadrangle, Charleston, SC provides a demonstration of the utility that near real time data can have for managing shellfish resources. An intertidal oyster resource assessment from the mid I980's was compared to the most recent oyster survey completed in the summer of 1995. In addition, an historical intertidal oyster survey completed in the winter of 1890-91 was included in the compar- ison. Qualitative observations made up the bulk of the historical survey, which included a limited but useful map of the resource. A GIS comparison of the three surveys illustrated the dynamic nature of the resource over an extended period. Field data included physical descriptions of the oyster beds, water temperature and salinity, tidal stage, locations of private and commercial docks, as well as erosional banks. The pc-based GIS allows data to be pre- sented in single or multiple map layers so that specific effects may be examined exclusively as well as inclusively. Comparative analysis over time allows managers to discern trends or abrupt changes in resource status. Policy measures can be adopted and implemented more accurately based on data and man- agement tools that contrast historic and contemporary information. The benefits of this system extend to commercial and recreational users as well as resource managers. RELATIVE EFFECTS OF HARVEST AND DISEASE MOR- TALITY ON EASTERN OYSTER POPULATIONS IN DEL- AWARE BAY. Stephen R. Fegley, Maine Maritime Academy, Castine, ME 04420; Susan E. Ford, John N. Kraeuter, and Harold H. Haskin, Haskin Shellfish Research Laboratory. Rut- gers University. Port Norris. NJ 08349; David R. Jones, Grice Marine Biology Lab. College of Charleston. Charleston, SC 29412. The decline in size of many Eastern oyster (Crassoslreci vir- giriica) populations and their failure to recover to historical abun- dances are frequently attributed to excessive harvests and the pres- ence of oyster diseases. Few attempts have been made to estimate the relative importance of these mortality sources to each other and to other factors affecting oyster abundance. We used a 40-year data set that included hydrography, abundances of oyster life- history stages, stage-specific oyster mortality, harvest intensity. and MSX disease prevalence and intensity to determine which of these factors most influenced adult oyster abundance on natural oyster beds in Delaware Bay. For all oyster beds combined, the prevalence of MSX disease was significantly and negatively correlated to oyster abundance; harvest volume was not related to subsequent oyster abundance. A highly significant, multivariablc regression model containing only three factors (1 -yearling oyster abundance, 2-mean annual Dela- ware River flow, and 3-mean proportion of oysters infected with MSX disease in the spring) explained almost two-thirds of the variance in adult oyster abundances on oyster beds. The absence of a general, negative effect of harvesting on Delaware Bay oysters probably results from the conservative management program ex- isting in the New Jersey oyster fishery since the 1950's. EASTERN OYSTER STOCK ASSESSMENT IN MARY- LAND. Mark L. Homer, Mitchell Tarnowski, and Lisa Baylis, Maryland Department of Natural Resources, Piney Point Aqua- culture Center, PC Box 150, Piney Point, MD 20674. During the last 13 years Maryland's public oyster fishery has declined from a 50 year average of 2.5 million bushels per season to less than 0.2 million bushels. Although this decline was and is concurrent with severe and widespread epizootics of Denno and MSX other causal factors such as overharvesting and habitat loss have been suggested. The debate that arose from attempts to at- tribute causality to the declining fishery led to the development of an oyster stock assessment program by the Maryland Department of Natural Resources in 1989. The initial goal of this effort was to produce a statistically sound sampling procedure that would give unbiased estimates of oyster abundance, population structure, and habitat volume. This was accomplished in 1990 through the de- velopment of a randomly initiated, systematic sampling scheme using hydraulic patent tongs covering 1 m" of substrate. Since then over 86,000 acres of charted oyster bars have been surveyed, including over 15,000 acres originally surveyed in 1975) when harvests were still high) using a similar and compatible method- ology. The results from the patent tong-based surveys indicate that there has been moderate to substantial habitat loss in some low to moderate salinity areas, and that these areas have been overhar- vested in recent years. In most of the moderate to high salinity areas, however, neither of these two factors can account for the severe harvest declines. Many areas will support oyster popula- tions as large (numerically) as those recorded in 1975 although severe truncation of size class structures and substantial decreases in the live to dead ratio of oysters have occurred. In addition, many of the moderate to high salinity areas have not been com- mercially exploited during the last 5 to 9 years and habitat loss in higher salinity zones has been minimal since 1975. Interestingly, the period from 1990 to the present has seen remarkable fluctua- tions in parasite severity, oyster recruitment, and freshwater dis- 522 Abstracl. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore, Maryland charge into the Chesapeake Bay. These fluctuations combined with the stock assessment data suggest that the primary factor in the decline of Maryland's oyster harvest has been parasite-related mortality. MORPHOLOGICAL DIFFERENTIATION OF THE FRINGING AND PATCH OYSTER REEF TYPES IN CHES- APEAKE BAY: A COMPARATIVE EVALUATION. G. F. Smith and K. N. Greenhawk, Cooperative Oxford Laboratory. Maryland Department of Natural Resources, Oxford. MD. Morphological and bathymetric examination of oyster reefs in Maryland Chesapeake Bay waters indicates that reefs may be seg- regated into two principal classifications: fringing reefs and patch reefs. Although gradation between the two reef types is common, morphological differences between these reef types can help ex- plain patterns of variation in historic cultch loss on charged oyster bars. Three dimensional analysis utilizing GIS technology has al- lowed for a general comparison of habitat loss from the turn of the century until the mid 1970's. A three dmiensional integration of bathymetry, historic charged oyster bar boundaries, and recent bottom composition surveys has identified principal causes of hab- itat loss for selected sites. Prmcipal causes of habitat loss due to sedimentation may be inferred to result from local processes rather than bay wide general siltation. MOLLUSCAN INVENTORY OF MARYLAND'S COASTAL BAYS. Mitchell L. Tarnowski.* Mark L. Homer, Lisa Baylis, and Robert Bussell, Maryland Department of Natural Resources. 580 Taylor Ave., C-2, Annapolis, MD 21401. MDNR"s Shellfish Program is m the third year of a compre- hensive effort to inventory the molluscan fauna of Maryland's coastal bays. Intended to establish baselme values for future man- agement needs, both commercially important molluscs and eco- logically valuable species have been targeted. To date approxi- mately 1700 stations have been sampled using five different col- lection methods. Nearly 50.000 individuals comprising 52 mollusc species have been collected. Two of these species represent range extensions and another 12 had not been reported in previously published accounts of the coastal bays. The first phase of the inventory was conducted in Chincoteague Bay, the largest and least environmentally impacted of the old coastal bays. Hard clam populations are 25'7f of the estimates made a quarter century ago. when hydraulic dredges were first introduced to the fishery. A juvenile hard clam survey showed recruitment to be substantially lower than in other regions. Severe predation. particularly by blue crabs, aggravated by the disappear- ance of oyster shell as a protective cover, may be a primary factor in limiting recruitment. The once highly prized Chincoteague oys- ter no longer inhabits the subtidal bars of the bay. having suc- cumbed to diseases, intense predation and competition for sub- strate. To a large extent the bars themselves have been buried by sediment or smothered by fouling organisms, greatly reducing this ecologically important habitat. Relic populations of oysters still exist intertidally, although the ribbed mussel Ceiikensia demissa is the dominant species in this zone. The inventory of ecologically valuable molluscs generated information on population and com- munity parameters such as species composition, distribution, abundance, size structure, and habitat characterization. Among the findings was the elucidation of ecological communities and func- tions of Chincoteague molluscs, the positive effect of detritus de- rived from salt marshes and seagrass meadows on molluscan abun- dance, and the importance of shell cover to species diversity. Also in Chincoteague Bay. an experiment is examining the growth and survivorship of hatchery reared, disease free oysters suspended in the water column. The inventory has been expanded into the other coastal bays, affording a synoptic view of the entire lagoonal system and allow- ing comparisons between relatively unimpacted and more de- graded ecosystems. Another round of sampling in Chincoteague Bay will give some idea of the temporal variability of population and community parameters. WATER QUALITY AND GOVERNMENT REGULATION CONTROL OF VIBRIO VULNIFICUS GROWTH TO RE- DUCE RISK IN SHELLFISH CONSUMPTION. Paul G. Co- mar,* National Marine Fisheries Service, Charleston Laboratory, Charleston, SC 29412. Vibrio vulnificus is a naturally-occurring bacterium present at elevated levels in estuarine waters and bivalve shellfish during warmer months. From 1992 through 1995. there has been an av- erage of 22 illnesses resulting in 10 deaths caused by V. vulnificus each year in the United States in raw shellfish consumers with several known pre-existing medical conditions. Nearly all of these shellfish illnesses have been linked to consumption of raw oysters from the Gulf of Mexico. Education of groups at risk for the disease and product advisory labeling at retail are risk reduction measures under development and implementation by regulatory agencies and the shellfish in- dustry through the coordination of the Interstate Shellfish Sanita- tion Conference (ISSC). In 1996. a new approach will be imple- mented to limit the increase in V. vulnificus levels in shellfish post-harvest and thereby reduce the risk of illness. It will require more rapid refrigeration of shellfish during warm weather months in states whose shellfish have been implicated in two or more V. vulnificus illnesses. This report describes the design of a cooperative analytical investigation to be conducted in 1996 to assess the control in V. vulnificus levels afforded by the new refrigeration requirements and how these data will be used with other infonnation in deter- minina the effectiveness of controls. National Shellfisheries Association. Baltimore. Maryland Abstract, 1996 Annual Meeting. April I'l-lS. 1996 523 ASSESSING WATER QUALITY: NEW DIRECTIONS. Eliz- abeth Fellows.* EPA Office of Water. 4503F. 401 M Street. Washington. DC 20460. Two major new activities will help the public and water man- agers understand water quality and set management priorities. The first is implementation of a nationwide strategy to improve water quality monitoring. The strategy was developed by the In- tergovernmental Task Force on Monitoring Water Quality (ITFM), a Federal/State consortium with an advisory committee of local and private experts. The strategy addresses nationwide mon- itoring design and collaboration, watershed and ecosystem com- ponents, environmental indicators, comparable monitoring meth- ods, quality assurance and control, assessment and reporting, and specific monitoring tools. The other activity is the first national water environmental indicators report that characterizes the na- tion's waters and how well we are meeting the goals of the Clean Water Act. The indicators measure how well the nation is doing to achieve goals of public and ecosystem health, attainment of water uses such as fishing and swimming, improvement of ambient con- ditions, and prevention or reduction of pollutant loadings and other stressors. Two of the indicators concerns shellfish consumption and the condition of shellfish beds. The indicators will depend upon and employ a wide range of data providers and users such as the shellfish management industry. TARGETING STRATEGIES FOR SHELLFISH RESTORA- TION IN THE GULF OF MEXICO: RESULTS OF A RE- GIONAL STRATEGIC ASSESSMENT PROCESS. Paul Or- lando, John Klein, Daniel Farrow. Anthony Pait, Dorothy Le- onard, and Jamison Higgins. Office of Ocean Resources Conservation and Assessment (N/ORCAl ). National Oceanic and Atmospheric Administration, 1305 East- West Hwy., Silver Spring. MD 20910. NOAA's Strategic Environmental Assessments Division (SEA), in partnership with EPA's Gulf of Mexico Program (GMP). has been conducting a strategic assessment to identify and geographically-target management strategies to meet the GMP's Shellfish Challenge, which is to "increase Gulf shellfish beds available for safe harvesting by 10 percent". The assessment pro- cess brings together key data sets characterizing shellfish bed clo- sure problems with state and regional experts to produce an inte- grated plan to meet the Shellfish Challenge. Since July 1994, there have been two regional workshops, numerous state visits, com- pletion of the 1995 National Shellfish Register, and the compila- tion of an extensive data base and mapping system that describes the classified shellfish harvest areas, pollution sources, estuarine salinity, and cultch planting activities. Nearly 30 strategies have been developed to reduce fecal coliform bacteria loadings, en- hance shellfish habitat enhancement, safeguard public health, and increase resource abundance. These strategies were prioritized and geographically-targeted to the 50 estuarine watersheds in the Gulf region. One of the highest-rated strategies was an expansion of cultch planting activities. This presentation will discuss how. through the strategic assessment process, areas suitable for shell planting were identified based on salinity requirements and resource productivity estimates in approved harvest areas (i.e., acceptable water qual- ity). The next step is to evaluate the feasibility of implementing several restoration strategies in one or more "good candidate" estuarine watersheds before proceeding with full-scale restoration. THE EFFECTS OF URBANIZATION ON THE AMERICAN OYSTER. CRASSOSTREA VIRGINICA (GMELIN). G. I. Scott,* M. H. Fulton, E. D. Strozier, P. B. Key, and J. W. Daugomah, US National Marine Fisheries Service, Southeast Fisheries Science Center. Charleston Laboratory, Charleston, SC; D. Porter, Marine Science Program University of South Carolina, Columbia, SC; S. Strozier. School of Public Health. University of South Carolina, Columbia, SC. Rapid development of coastal areas of the southeastern US has resulted in significant alterations of upland terrestrial habitats ad- jacent to sensitive estuarine salt marsh ecosystems in the south- eastern US. Most remaining coastal development in the southeast- em US will be residential and tourism/service related industries rather than industrial development and will occur around the >300 small high salinity tidal creeks and estuaries found in the region. These alterations may result in potential impacts to living re- sources within estuaries, including molluscan shellfish such as the American oyster, Crassostrea virginica. The Urbanization in Southeastern Estuarine Systems (USES) study has addressed impacts of coastal development on adjacent small, high salinity estuaries of the southeastern US by comparing Murrells Inlet (MI), a highly urbanized estuary with pristine North Inlet (NI). A total of 60 monitoring stations were sampled in both estuaries. Surface waters, sediments and oysters (Crassostrea vir- ginica) were monitored for fecal coliform bacteria densities; sero- typed to individual bacterial species; analyzed for trace metals, polycyclic aromatic hydrocarbons (PAHs). pesticides and poly- chlorinated biphenyls (PCBs) to characterize chemical contami- nant inputs; and adult oysters were monitored for survival, con- dition index, gonadal index and juvenile spat settlement. Geo- graphical Information Processing (GIP) was conducted on multiple data layers to indicate geographic regions where multiple contam- inant interactions had occurred. One of the more significant effects from urbanization study was the increased closure of shellfish harvesting waters due to in- creased inputs of fecal coliform bacteria. More than 67% of the sampling sites in Ml exceedede the SA water classification fecal coliform standard versus 35% in NI. Fecal coliform monitoring of shellfish meats indicated that >50% of stations in each estuary exceeded the Interstate Shellfish Sanitation Conference Depura- tion Meat Standard. Mortality rates among adult and juvenile oys- ters was much higher in Ml than NI. and the pattern of spat settlement was different. GIP aqnalysis indicated areas where mul- 524 Abstract. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association, Baltimore, Maryland tiple contaminant interaction occurred and where coastal ecosys- tem health was adversely affected. SYNOPSIS OF FDA RESEARCH RELATED TO WATER QUALITY. William D. Watkins,* HFS-417, Office of Seafood, U.S. FDA, 200 -C Street. S.W.. Washington. D.C. 20204. FDA research involving water quality focuses primarily on issues involving molluscan bivalve shellfish and the National Shellfish Sanitation Program. The goal of this research is to enable the prevention of various health hazards sometimes associated with the consumption of raw shellfish, and to provide sound sci- entific data for regulatory and compliance decisions. Research involving traditional and alternative indicators of fe- cal contamination demonstrate the utility and significance of using an array of microbial indicators. This array variously includes total and fecal coliforms. Escherichia coli. enterococci. Clostridium perfringens. and male-specific bacteriophage. Studies have pro- vided valuable information on shellfish depuration, wastewater disinfection, microbial survival, assessments of sanitation in ma- rine environments, and shellfish-fome illness outbreaks. Other studies on densities of the naturally occurring, opportunistic patho- gen. Vibrio vulnificus, show the importance of storage tempera- tures (and times) and elucidate the effects of applying various intervening measures. Water quality research related to shellfish toxins suggest that phytoplankton monitoring may provide reliable indications of impending toxic algal blooms. Several important issues related to molluscan shellfish remain unresolved. Future water quality research will need to address the development of new criteria for effective closure zones around point sources, re-opening criteria for offshore areas, and a more human-specific indicator of fecal contamination. POSTER SESSION INCIDENCE OF FOULING AT TWO MARICULTURE SITES IN BON SECOUR BAY, ALABAMA. Susan B. Atha- nas* and David B. Rouse, Department of Fisheries and Allied Aquacultures, Auburn University, Auburn, AL 36849. Private oyster producers in Bon Secour Bay, Alabama have recently adopted off bottom aquaculture practices. Barnacle foul- ing has increased tremendously, increasing labor cost for market preparation. A survey was conducted from July 1994 to June 1995 to determine seasonality of barnacle spawning in order to discover possible prevention techniques. Location withm the bay and within the water column were also examined. Balanus eberneus was determined to be the only barnacle species fouling oysters. Most barnacle sets occurred from January to May with peak foul- ing observed in March. Little differences were observed in loca- tion or depth. Possible management strategies might be to stock young oysters in the bay after June, allowing the accumulation of siii and bryazoans which will discourage later barnacle fouling. CAN A SPECIES BE FOULED" INTO EXTINCTION? ZE- BRA MUSSELS VS NATIVE BIVALVES IN THE UPPER MISSISSIPPI RIVER. P. Baker* and D. J. Hornbach, Maca lester College. St. Paul, MN 55105. Zebra mussels (Dreissena) are known to locally eradicate na- tive clam (Unionaeea) populations, by heavily fouling the exposed shells. Native bivalves thus far impacted have been protected from extinction by their large ranges, which include waters unsuitable for Dreissena. As Dreissena spreads, however, it has encountered unionaceans with restricted ranges. Our research examined the threat posed to one such species, the endangered Lampsilis hig- ginsi. All but two of the reproducing populations of L. higginsi are in the mainstem Mississippi River. The remaining populations are in the lacustrme downstream region of the St. Croix River (MN & WI). Dreissena have reached high densities on all Mississippi habitats, and are in the process of invading the lower St. Croix. Water chemistry and the long water residence time in the lower St. Croix are favorable for Dreissena reproduction. If Dreissena in- vades as predicted, and affects unionaceans as it has elsewhere, uncommon species such as L. higginsi will probably be reduced to densities below the minimum required for successful spawning. Extinction could follow in the time it takes for the remaining L. higginsi to die of old age. Transplanting or cultunng L. higginsi may preserve this species. PADDLES OR SIEVES: TESTING THE MECHANISM OF PARTICLE RETENTION IN BIVALVES. Kerri M. Bent- kowski* and J. Evan Ward, Department of Biological Sciences. Salisbury State University. Salisbury. MD 21801; Roger I. E. Newell, Horn Point Environmental Laboratory. University of Maryland. Cambridge, MD 21613. The mechanism of particle capture in suspension-feeding bi- valves is controversial and poorly understood. Despite the tradi- tional view that the gill and its associative rows of laterofrontal cilia or cirri physically trap particles m a manner similar to a filter, recent insights into the physical forces that interact at small size scales show an inconsistency in the paradigm. For example, vis- cous forces of the water tend to over-nde inertial forces at the size scales of the cilia. In order to test hypotheses concerning the forces that mediate particle capture, we manipulated the kinematic viscosity of the fluid by changing water temperature. Retention efficiency was measured for particles I p.m and 10 |j,m in diameter with a Coulter Multisizer. Previous workers have found that retention efficiency of 1 |xm particles in some planktonic suspension-feeders is medi- ated by fluid viscosity. In contrast, our preliminary data with M. edulis indicate that retention efficiency of 1 |j.m particles is inde- pendent of kinematic viscosity even when viscosity increases by 817r. Our results support the hypothesis that laterofrontal cirri function more like paddles than like sieves. Implications of our results to feeding mechanisms in bivalves will be discussed. National Shellfisheries Association. Baltimore. Maryland Ahstnui. 1996 Annual Meeting. April 14-18. 1996 525 SEASONAL CYCLE OF HAPLOSPORIDIUM NELSONI (MSX) IN INTERTIDAL OYSTERS, CRASSOSTREA VIR- GINICA, IN SOUTH CAROLINA. M. Yvonne Bobo* and Donnia Richardson, South Carolina Department of Natural Re- sources (SCDNR). Marine Resources Research Institute. Charles- ton, SC 29412; Thomas C. Cheng, Shellfish Research Institute. Charleston. SC 29412: Elizabeth McGovern and Loren Coen, SCDNR. Marine Resources Research Institute. Charleston. SC 29412. Little is known about the seasonal cycle of the oyster pathogen Haplosporidium nelsoni (MSX) in the southeastern United States. South Carolina oysters are predominantly intertidal (>95%) and occupy a very different habitat from their northern counterparts which are predominantly subtidal. MSX disease was first reported in SC in 1993 in oysters collected from Charleston Harbor and subsequently observed in oysters from other SC coastal sites. Dur- ing summer, 1994, an extensive survey was conducted to deter- mine the prevalence of MSX over a larger geographic area. Twenty-one stations were sampled (n = 925 oysters examined). MSX was present in oysters from 10 of the 21 stations surveyed (48%), with prevalences as high as 28%. In June 1994, monthly sampling of a Charleston Harbor site was initiated. In September 1994, as part of a long-term intertidal oyster ecosystem study, two additional sites were added, one at a degraded marina and the other in a pristine tidal creek system surrounded by marsh and mudflats. Monthly sampling (n = 25) from each of these three sites has continued since June 1994. and has revealed that the highest prev- alence and intensity occurred during the Fall. MSX was present at all three sites, with maximum prevalence ranging from 28 to 42% and intensities ranging from rare to heavy. All diagnoses were by examination of representative histological sections. Concurrent physical environmental data (DO, temperature, salinity) were col- lected at the sites. Mortalities due to MSX were not apparent. EFFECTS OF PERKINSUS MARINUS ON CULTURED MO- BILE BAY OYSTERS. Guy W. Brunt and Yolanda J. Brady, Department of Fisheries and Allied Aquacultures, Auburn Univer- sity, Auburn, AL 36849-5419. A protozoan parasite, Perkinsus mannus. is both common and widespread in the Gulf of Mexico and has been responsible for severe infection and mortality of eastern oysters, Crassostrea vir- ginica. This study investigated the effects oiP. mahnus on oysters being cultured in a suspended bag system in the Bon Secour area of eastern Mobile Bay, Alabama. Determinations of infection prevalence and intensity as well as related regulating factors were made, as were determinations of the effects of P. mahnus infec- tion on oyster digestive diverticula and vesicular connective tissue. Also, differences between first and second year age class oysters and differences between oysters cultured near the top and bottom of the water column were assessed. P. mahnus infection prevalences were high (usually 85% or above) throughout the study period. Infection intensities were gen- erally light, however, and were not ususally at levels associated with severe oyster mortality. Higher infection prevalences and intensities were associated with the warmer months, though salin- ity was concluded to be the primary regulating factor. Condition of oyster digestive diverticula and vesicular connective tissue de- creased as infection intensity increased, although other factors probably exerted some influence on oyster condition as well. No consistent differences were found between oyster groups (first and second year, high and low in water column), with the exception that second year oysters had more atrophic digestive diverticula than first year oysters. PREDATOR- AND PREY-DIFFERENTIATED REPAIR FREQUENCIES IN THE MOON SNAILS EUSPIRA HEROS AND NEVERITA DUPLICATA VERSUS THE WHELKS BUSYCON CARICA AND BUSYCON CANALICULATUM FROM CAPE MAY COUNTY, NEW JERSEY. Gregory P. DietL Dcpt. of Geological and Marine Sciences. Rider University. Lawrenceville, New Jersey 08648. More than 1300 specimens combined from the moon snails Euspira heros and Neverita duplicala. and the whelks Busycon cahca and Busycon canaliculatum, were collected from Hereford Inlet and Great Egg Harbor in Cape May County, New Jersey. Body whorl diameter and apertural lip thickness were measured, and the number of sublethal scars per final whorl counted, for each specimen. Although the mean number of repairs per specimen was different among the four species (ANOVA, p = .0001), the av- erage was comparable for the two moon snails, namely 1 . 1 and 0.9 for A', duplicala and £. heros. and identical (6.3) for B. cahca and B. canaliculatum. Repair frequencies per specimen ranged from 0 to 13 for both whelk species, 0 to 12 for/V. duplicata, and 0 to 7 for E. heros. Only 4% and 3% of B. canaliculatum and B. cahca, respectively, lack repairs, whereas 48% and 57% of £. heros and A'., duplicata. respectively, lack repairs. Repair frequencies are strongly correlated with both body whorl diameter and apertural lip thickness for both whelks and A', duplicala. but not for E. heros from the Hereford Inlet locality. The greater mean frequency of repairs in B. canaliculatum and B. cahca relative to the two moon snails is attributed to shell breakage from predation by crabs on whelks combined with ap- ertural lip fracture during attempts by whelks to wedge apart the valves of their bivalve prey. In contrast, manipulation of shelled prey by moon snails in preparation for drilling doesn't induce apertural fracture. Consequently, less than 5% of whelk shells show no damage in contrast to the more than 50% of moon snails lacking repairs. Surprisingly, the thinner whelk B. canaliculatum and moon snail E. heros do not have number of repairs-frequency distributions different from their thicker lipped relatives B. carica and A', duplicata. respectively. (Kolmogorov-Smimov test; /? = .20 between whelks; p = .28 between moon snails). Repaired fractures on the apertural lips of the largest whelks and moon snails indicate a lack of a "size refuge" from sublethal predation. 526 Absiraa. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore, Maryland SPATIAL PATTERNS OF INTERTIDAL OYSTER REEFS IN THE CANAVERAL NATIONAL SEASHORE, FLOR- IDA. Ray Grizzle,* Randall Environmental Studies Center. Tay- lor University. Upland. IN 46989; Mike Castagna, Virginia In- stitute of Marine Science. College ot William and Mary, Glouces- ter Point. VA 23062. The Canaveral National Seashore includes much of the north- em Mosquito Lagoon in northeastern Florida, and some of the few remaining large intertidal oyster (Crassostrea virginica) reefs along the Atlantic coast. We characterized intra-reef patterns at a scale of 1 m by sampling quadrats quarterly (July 1994-March 1995) along fixed transect lines on ten different reefs, and lagoon- wide (inter-reef) distribution patterns using low-altitude aerial photography (1:6.000 imagery, color IR film) in January 1995. Intra-reef patterns included a strong "edge effect" with sub- stantially greater spat settlement and oyster density (but no differ- ences in mean shell height) in a 2 to 3-m fringe around most reefs. Inter-reef patterns showed two strong trends; a lagoon-wide S to N increase in areal coverages by the reefs correlated with increasing tidal ranges; and multiple-reef patterns that suggested a relation- ship to tidal flow patterns. Tide range was also positively corre- lated with oyster and spat densities, but not oyster size. Though a complete analysis has not been done, reefs showed obvious com- plex spatial patterns relative to tidal channels. The largest reefs and many small reefs were oriented parallel to and/or along the edges of major tidal channels. In contrast, there were several clus- ters of moderate-sized reefs arranged in dendritic patterns associ- ated with multiple tidal channels. Both intra- and inter-reef pat- terns can be explained by existing hydrodynamica! models con- cerned with water flow and food fluxes. SERIAL DEPLETION IN MARINE INVERTEBRATE DIV- ING FISHERIES. Peter L. Haaker,* California Department of Fish and Game, Long Beach, CA 90802; Gary E. Davis, National Biological Service, Ventura, CA 93001; Ian K. Taniguchi, Cal- ifornia Department of Fish and Game, Long Beach. CA 90802. California's diving fisheries relied primarily on five abalones. one sea urchin, and a few other invertebrates in the 20th century. Improvements in technology, e.g., fast boats, electronic naviga- tional and survey equipment, SCUBA, and thermal protection for divers, increased diver efficiency and provided access to remote territories and deep habitats. As populations of the most accessible and valuable species declined, divers switched to less accessible and lower valued species. Red abalone dominated landings prior to Worid War II. followed by pink abalone in the 1950s and 1960s. Green and white abalones contributed briefly in the mid-1970s, before harvest shifted to intertidal black abalone and red sea ur- chins. Current abalone landings are —10% of historic levels. When urchin landings in southern California began to wane in the c:irly 1980s, fishing effort moved to new territories in northern Caiifomia. The shift from abalones to urchins required an order of magnitude increase in harvested biomass to sustain total fleet in- come, e.g., 2,000 tonnes of abalone yielded about the same in- come as 20,000 tonnes of sea urchin. This frontier approach to sustained fisheries continues today. The dive fishery is currently exploring more species, e.g., small purple sea urchin, wavy top turban snail, purple coral, and sea cucumber. As species were depleted, their harvest was not curtailed to allow recovery, consequently those populations remain at danger- ously low levels. When new fisheries were developed no entry restrictions were imposed, leading to overcapitalization and over fishing. As technological improvements increased the efficiency of harvest, management paradigms failed, and important marine resource stocks declined to the point of collapse. Even though serial depletion of marine resources may have been advantageous to the fishing industry in the 20th century, it is not sustainable. Restoration of public benefits derived from the productivity of coastal waters will now require expensive restoration of depleted populations and a loss of benefits for many decades while resto- ration is affected. FEDERALLY ENDANGERED FRESHWATER MUSSELS IN THE ST. CROIX RIVER: MICROHABITAT AND MUS- SEL COMMUNITY ASSOCIATIONS. D. J. Hornbach,* T. Deneka, and P. Baker, Dept. Biol , Macalester Coll., St. Paul, MN 55105. Recent studies indicate that macrohabitat rather than microhab- itat factors are better predictors of freshwater mussel community structure. However, microhabitat factors can be important in de- scribing local abundance and distribution of mussels. Unfortu- nately, for many rare mussels, factors responsible for controlling their abundance and distribution have not been studied. Our hy- pothesis was that endangered mussels are found in high quality mussel habitat rather than in peculiar niches and must be found in localized populations if they have been successful in maintain a viably reproducing population. We conducted this study in a Mississippi River tributary, the St. Croix River, which contains at least 38 mussel species includ- ing two federally endangered species; Lampsilis higginsi and Qua- drula fragosa. Using SCUBA. 494 0.25 m' quadrats were exam- ined to characterize the mussel community and microhabitat (sed- iment size, water depth and flow). We also conducted searches specifically for Q. fragosa and L. higginsi quantitatively sampling areas where specimens were found. The most dense and rich mussel communities were associated with specific substrate types in conjunction with particular water depth and flow regimes. Q. fragosa and L. higginsi were found in areas of rich and diverse mussel assemblages. Consequently these endangered species did not have peculiar niches, indicating that a community assessment technique may be helpful in endangered mussel management. National Shellfisheries Association, Baltimore, Maryland Abstract. 1996 Annual Meeting. April 14-18. 1996 527 THE EFFECTS OF LARVAL STOCKING DENSITY ON GROWTH AND SURVIVAL OF LABORATORY REARED SPISULA SOLIDISSIMA SIMIUS. Dorset H. Hurley* and Randal L. Walker, Shellfish Aquaculture Laboratory, University of Georgia, Marine Extension Service. 20 Ocean Science Circle, Savannah, GA 31411-1011. Growth and survival oi Spisiila solidissima similis (Say, 1822) larvae stocked at 10, 20. 30 and 50 larvae per ml were determined in a laboratory growth study to define the optimum stocking den- sity for culture of this subspecies. Twenty-four hour old larvae were stocked at the above densities within 500 ml flasks containing filtered seawater at 25 ppt and 20°C in a constant temperature-controlled room. All treatments received a daily food ration of 100.000 cells per ml of Tahitian strain Isoclinsis sp. with a complete water exchange per flask every two days. Three replicate flasks per treatment were subsampled (N = 5), on days 1, 5. 9, 15, and 27. No significant differences (p = 0.3539) in survival occurred between stocking den- sities at day 27 with percent survival ranging from 61% for the 10 larvae per ml to 32% for the 50 larvae per ml stocking treatments. Larval size ((i.m) was significantly different for all treatments on day 9 (p< 0.0001: 50, X = 88.4 < 20, x = 95.0 < 30. x = 101.8 < 10, x= 1 17.0) and day 27 (p< 0.0001; 50. x = 145.8 < 30. x = 175.6 < 20, X = 195.7 < 10, x = 263.0). A higher percentage of animals had undergone complete metamoiphosis at day 27 in the lower stocking density' treatment of 10 (87%) than in the higher stocking density treatments of 20, 30, and 50 larvae per ml (13%, 3.6%, and 0%, respectively). SALINITY EFFECTS ON INTRODUCED DREISSENID MUSSELS, V. S. Kennedy.* M. Aspleen, and T. Hall. Uni versity of Maryland System. Horn Point Environmental Labora- tory, Cambridge, MD 21613. We tested the effects of salinity on movement, byssal attach- ment, and mortality of smaller (<1.5 cm) and larger (>2 cm) zebra {Dreisseiia polymorpha) and quagga (D. bugensis) mussels acclimated to freshwater, and to 3 and 5 ppt (quagga mussels; QM) or to 4 and 6 ppt (zebra mussels; ZM). Freshwater-acclimated mus- sels demonstrated a decrease in movement after 48 h in salinities of 7-1- ppt (ZM) and 4+ ppt (QM), a decline in byssal attachment at 5-1- ppt (ZM) and 3+ ppt (QM). and increased mortality at 6 ppt (ZM) and 4 -I- ppt (QM). Acclimation to higher salinities at a rate of 1 ppt every 3 d raised the salinities at which movement and attach- ment were inhibited after 48 h for both mussel Sf)ecies. Zebra mussels survived exposure to 10 ppt with limited mortality (>30%); quagga mussels tolerated up to 8 ppt. Larger zebra and quagga mussels were less likely to move, attach, or survive than were smaller mussels in similar salinities. Additional experiments on the native dreissenid in Chesapeake Bay, Mytilopsis leucophaeata. found that movement, attachment, and mortality were minimally affected by salinities down to 2 ppt for individuals acclimated to 12 ppt. with limited mortality even in fresh water. Some zebra and quagga mussels spawned spontaneously in the experimental bowls during the salinity tolerance experiments. Em- bryos formed in salinities above about 4 ppt did not survive past the early stages of cell division, even if derived from mussels acclimated to higher salinities. Eggs spawned into fresh water by mussels acclimated to 3 to 6 ppt went unfertilized, ultimately enlarging, then disappearing. We conclude that zebra and quagga mussels can adjust to oligohaline conditions, that they could over- lap with M. leucophaeata. but that spawning in the upper estuary may not produce viable embryos. GENETICS AND SYSTEMATICS OF FRESHWATER MUS- SEL SPECIES: A TISSUE REPOSITORY. Tim L. King.* Mary E. Smith, Rita F. Villella. Priscilla I, Washington, and David A. Weller. Department of the Interior. National Biological Service — Leetown Science Center, Aquatic Ecology Laboratory, 1700 Leetown Road, Keameysville, WV 25430. Recognizing a lack of population genetics and phylogenetics information with respect to freshwater mussel conservation ef- forts, the National Biological Service — Leetown Science Center's Aquatic Ecology Laboratory has established a repository and as- sociated database to coordinate tissue samples collected for genet- ics and systematics research. The repository provides a centralized location for researchers to obtain properly catalogued and pre- served adductor muscle, mantle, foot, gill (including glochidia). and digestive gland tissue samples. All data generated for the repository are maintained in the PARADOX for Windows rela- tional database package. Collection information compiled for each specimen includes data, site name, site description, and habitat characteristics. Database content reports are generated and pro- vided to interested researchers. Currently the database contains in excess of 260 individuals representing 46 species inhabiting At- lantic slope and Interior Basin drainages. All researchers utilizing the repository are required to accommodate a standard numbering scheme to allow comparisons of the same individuals among di- verse studies and methodologies. Potentially, the repository would reduce the number of animals sacrificed and sampling time while providing comprehensive data to multiple researchers. A single collection of mussels can provide ecophenotypic, protein, DNA, and immunological information for species and population struc- ture delineation. This poster describes the development of the repos- itory, provides instructions for tissue contribution and retrieval, and presents data collection protocols, preservation methods, database structure, and a current report of the database contents. DISCONTINUITY IN THE GENETIC POPULATION STRUCTURE OF THE GREEN FLOATER LASMIGONA SUBVIRIDIS. Tim L. King,* Rita F. Villella. Mary E. Smith, and Michael S. Eackles. Department of the Interior, National Biological Service — Leetown Science Center, Aquatic Ecology Laboratory. 1700 Leetown Road. Keameysville, WV 25430. Modern molecular genetic techniques have revealed population genetic structure and phylogenetic associations within and among many rare taxonomic groups. However, little genetic-based pop- 528 Abslraa. 1996 Annual Meeting. April 14-18, 1996 National Shellfisheries Association, Baltimore, Maryland ulation or phylogenetic information is available for most freshwa- ter mussel species. To address the need for mussel genetics re- search, we have evaluated techniques to identify and assess ge- netic variability in selected ribosomal and mitochondrial (mtDNA) DNAs among geographic populations of the green floater Lasmig- ona subviriciis. A total of 43 L. siibvindis were sampled from nine geographic populations representing five rivers draining the At- lantic slope (Susquehanna, 2 sites; Potomac. 2 sites: Rappahan- nock; James; and Neuse Rivers) and the New River (2 sites) an Interior Basin drainage. Each mussel was subjected to polymerase chain reaction amplification and restrictionase digests of the first internal transcribed spacer region (lTS-1) of nuclear ribosomal DNA and the cytochrome oxidase I (COl) region of mitochondrial DNA. Diagnostic genetic differentiation was observed in both ri- bosomal and mitochondrial DNA among geographic populations of L. siibviridis. The 570 base-pair (bp) ITS-1 fragment, digested with the enzyme Ddc 1, produced a diagnostic restriction site poly- morphism between the northern populations (Susquehanna and Potomac Rivers) and the remaining populations. The northern populations also exhibited a diagnostic restriction site polymor- phism in the 710 bp COl region of mtDNA when digested with the enzyme Cfo I. Preliminary results suggest the presence of a lati- tudinal discontinuity in the population structure of L. siibviridis. and possibly the absence of gene exchange. No differentiation among the Atlantic slope and Interior Basin populations has been observed. EVALUATION OF TAG TYPES AND ADHESIVES FOR MARKING FRESHWATER MUSSELS. David P. Lemarie, David R. Smith, Rita F. Villella, and David A. Weller. National Biological Service — Leetown Science Center, Aquatic Ecology Laboratory, 1700 Leetown Road, Kearneysville, WV 25430. External identification of individual mussels is highly desirable for following passive and active movements, population studies, and labeling for studies of growth, reproduction, genetics, and physiology. Ideally, tags must be easy to apply, inexpensive, and provide excellent long term legibility and retention. In this study we evaluated three varieties of tags (Northwest Marine Technology Visual Implant Tag, Floy Fingerling Tag, and Hallprint Shellfish Tag), two types of adhesives (3M two-part epoxy and Krazy Glue cyanoacrylate). and four bonding times before immersion in water (2, 5, 10, and 15 min). Tags were applied to shells of dead animals. Tag/glue combinations showing good initial legibility after complete curing of the adhesive were further tested under natural conditions in a shallow stream and in a standard gem tumbler containing coarse metal shavings. This poster provides an illustrated summary of the advantages and disadvantages of each of the tag types and adhesives tested. Preliminary results suggest that the best combination is a flexible polyethylene shellfish tag bonded to the shell with cyanoacrylate. Cyanoacrylate can be immersed in water in as little as two minutes after application. SPECIES SPECIFICITY AND EFFECT OF PH ON THE RE- SPONSE OF FRESHWATER MUSSEL JUVENILES TO ACUTE COPPER TOXICITY. A. D. McKinney, Tennessee Wildlife Resources Agency, Nashville, TN; R. G. Hudson, Pres- byterian College, Clinton, S.C; Margaret L. Barfleld, Arkansas State University. Jonesboro, AK. Unerbackia imbecilis is the only freshwater mussel species whose juveniles have routinely been used for toxicity testing. This species is somewhat ubiquitous, occurring in a variety of habitats. A review of literature showing that pH affects the toxicity of certain metal ions in nonmolluscan species is given. To determine if this is true also with mussel juveniles, a range of copper ion concentrations was tested on U . imbecilis ]\x\tm\ts at two different pH levels. Furthermore, to show whether this species is represen- tative of a flowing water species, Ellipilio angusiaia juveniles were also tested at the more neutral pH range. Tests were made with four repetitions of 10 juveniles/repetition in each test group using moderately hard water with 800 mg/l silt and the addition of bloomed plankton, imitating their natural environment as much as possible. Copper concentrations ranged from 1^ ppm, with a control. Results show that the lower pH values caused the sensi- tivity of the juveniles to more than double, with a LC50 value of 1.28 ppm in the lower pH group and 2.75 ppm in the higher pH group. Comparison of the higher pH group with juveniles of E. angustata revealed that the two had identical LC50 values. The immediate significance of this preliminary study is to emphasize the consideration of pH when conducting and comparing toxicity studies involving juvenile mussels. Furthermore, the increased stress on populations in acidified water is obvious, and standards regulating known pollutants should be qualified by a statement of acceptable pH range. Finally, juveniles of U. imbecilis were rep- resentative of this other stream dwelling mussel as far as copper sensitivity is concerned. FATE OF POTENTIAL BACTERIAL CONTAMINANTS AS A FUNCTION OF CONTACT SURFACE IN SHELLFISH WET HOLDING TANKS. Carter R. NewelL* Great Eastern Mussel Farms, Inc., Tenants Harbor, ME 0486; Bohdan M. Slabyj, Department of Food Science, University of Maine, Orono, ME 04469. Perceived risks associated with the use of masonry surfaces in shellfish holding tanks and associated reservoirs have led to the enforcement of "food contact surface" requirements in those sys- tems. In preparation for the 1995 ISSC meeting, a study was performed using both porous (cement) and non-porous (fiberglass) mussel (Mytilus edulis) miniature wet holding tanks (microcosms). The incoming water was seeded with Escherchio coli. Enterococ- cusfaecium. Listeria innocua and a non-pathogenic Vibrio isolate. Influent and effluent water, tank sediments and mussels were sam- pled from 0-72 hours. Mussel lots were changed daily, and tanks were steamed cleaned between lots mirroring normal production conditions. In both types of tanks (concrete and fiberglass) all National Shellfisheries Association. Baltimore. Maryland Abstract. 1996 Annual Meeting. April 14-18. 1996 529 seeded bacteria disappeared with time and were essentially re- moved with routine cleaning. The experiments should be repeated for confirmation. Nonetheless, they indicate the beneficial role of the natural seawater flora in outcompeting potential pathogens in seawater from approved shellfish growing areas. RFLP ANALYSIS OF GENETIC DIVERSITY IN A SIBE- RI.4N POPULATION OF THE JAPANESE SCALLOP {PA- TISOPECTEN YESSOE.VSIS). Elizabeth A. Orbacz.* Ami E. Wilbur. Jeffrey R. Wakefield, and Patrick M. Gaffney, Col- lege of Marine Studies. University of Delaware. Lewes, DE 19958. Restriction fragment length polymorphism (RFLP) analysis was used to compare the genetic diversity of a Siberian population of the Japanese scallop {Palinopecten yessoensis) with populations previously examined by Boulding et al. (1993. Can. J. Fish. Aquat. Sci. 50: 1 147-1 157) including a small hatchery population in British Columbia and two wild populations from Mutsu Bay (Aomori) and Uchiura Bay (Hokkaido) in Japan. The polymerase chain reaction (PCR) was used to selectively amplify three coding regions of the mitochondrial genome in 20 individuals from Peter the Great Bay (Primorye Region. Russia). These regions included ( 1 ) most of ATP synthetase subunit 6 and cytochrome c oxidase subunit 3 (1.5 kb).(2) tRNA for threonine (1.1 kb). and (3) part of cytochrome b apoenzyme ( 1 .4 kb). Digestion of the PCR prod- ucts with 1 1 restriction enzymes revealed 22 polymorphic sites and 19 distinct composite haplotypes. Haplotype diversity and within- population nucleotide diversity were high in the Siberian popula- tion (0.98 and 0.015 respectively). Both of these estimates of genetic diversity are much greater than those calculated by Boul- ding et al. (1993) for the hatchery (haplotype diversity = 0.53, nucleotide diversity = 0.0012) and Japanese populations (mean haplotype diversity = 0.72. mean nucleotide diversity = 0.0017). MANAGEMENT STRATEGIES FOR FOULING CONTROL IN ALABAMA OYSTER CULTURE. F. Scott Rikard and Richard K. Wallace, Auburn University Marine Extension and Research Center. Mobile. AL; Christopher L. Nelson, Bon Secour Fisheries, inc.. Bon Secour. AL. Fouling by marine organisms is a major impediment to the development of inshore mariculture. Fouling control methods for off bottom oyster culture were analyzed experimentally over a two year period for effects on fouling, oyster growth, oyster condition and oyster survival. Oysters were held in plastic mesh bags at- tached to a belt system suspended in the water column. The first year study focused on pressure washing treatments at 2, 4. and 8 week intervals, and biological control treatments using blue crabs, hermit crabs and stone crabs, and a control receiving no washing or animals. Frequentlyt washed oysters (2 and 4 week intervals) had significantly less fouling than the 8 week wash interval or the unwashed control but were significantly smaller and suffered greater mortality. Stone crabs showed the most potential for bio- logical fouling control but also appeared to prey significantly on the oysters. Based on the first years results, methods were refined and a second experiment designed. Treatments analyzed were a 6 week was interval, a bag change treatment, biological treatments using larger blue crabs, and a control. There were no significant differences in fouling, mortality and condition among all treat- ments at the time of harvest. Some significant differences in the fouling index between 6 week wash interval and other treatments were seen during peak fouling times. There was a significant dif- ference in growth between the bag change treatment and the con- trol at the time of harvest. Current management suggestions are to pressure wash bags at a 6 week or greater interval and also target washing to coicide with peak settlement of fouling organisms. SENSORY PHYSIOLOGY OF GLOCHIDIA LARVAE OF THE FRESHWATER MUSSELS UTTERBACKIA IMBECIL- LIS AND MEGALONAIS NERVOSA. Melanie K. Shadoan* and Ronald V. Dimock. Jr.. Department of Biology. Wake For- est University, Winston-Salem. NC. The stimuli involved in the attachment of larvae of freshwater mussels to fish hosts are almost completely unknown, with a few observations suggesting fish mucus or body fluids as an important cue. The responses of glochidia of Utterbackia imbecillis and Megakmais nervosa to chemical and mechanical stimulation were monitored. Larvae with hooked shells (U . imbecillis) typically attach to fins and opercula. while bookless larvae (M. nen^osa) attach to gills. The larvae were exposed to fish epithelial mucus (blue gill. bass, goldfish) and to mucus that was fractioned by ultrafiltration (fractions; >IOKD. <10 >3KD, <3KD). Putative components of fish mucus including sialic acid, galactose, man- nose, and free amino acids (pH 6.5) also were tested. In addition, larvae were mechanically stimulated (stroked until tonic closure) with a 30 jj.m glass micropipet in a micromanipulator and also mechanically stimulated in the presence of selected amino acids to determine if a synergistic effect between chemical and mechanical stimulation occurred. When exposed to fish epithelial mucus, all larvae of both species of mussels experienced tonic closure; how- ever, the duration of closure was longer for U. imbecillis. Glochidia responded only to the <3KD fraction of fish mucus. Of the non-amino acids tested, only sialic acid induced rhythmic ad- ductions, with M. nervosa showing a greater response. M. nerx'osa was also more sensitive than U. imbecillis to all tested amino acids at 10~"M. responding with more rhythmic adductions and a higher percentage of larvae undergoing tonic closure. Mechanical stimulation induced tonic closure in both species, but M. nenosa required more stimuli before it closed. Synergism between me- chanical and chemical stimuli was evident from a significant in- crease in the duration of tonic closure by larvae of both species. 530 Abstracl. 1996 Annual Meeting. April 14-18. 1996 National Shellfisheries Association. Baltimore. Maryland THE CULTURE VARIABILITY OF MONOCHRYSIS LVTHERl AS AN ADVANTAGE FOR SHELLFISH CUL- TURE. Lioudmila V. Spektorova, Harbor Branch Oceano graphic Institution. 5600 Old Dixie Highway. Fort Pierce, FL 34946. During different developmental stages, mollusks need algae of various chemical compositions. Aquaculturists require better knowledge of conditions that ensure a biomass yield rich in pro- tein, lipids or with a high HUFA level. From this point of view. Monochnsis lutlieri is attractive, because it has a flexible metab- olism. M. lutheri contains high level of lipids and both essential fatty acid 20:5w3 and 22:6w3. in similar proportions to oyster tissue. Microalgae experiments were made outdoors in a closed tubu- lar photobioreactor of 160 I volume capacity on the Black Sea Experimental Station between May and October. The daily radiant energy averaged 1 32 W m " " in May. 171 W m " " in July and 80 W m " in October. Nutrient concentrations were maintained at 230-250 mg N T '. 50-70 mg P r ' at a pH 7.0-7.7. The optimal temperature for this strain was 28°. We tested the role of three factors in influencing the chemical composition of M. liiiheh: season, amount of nitrogen, source of nitrogen. Most favourable conditions were in July, when the culture reached a density of 400-600 10^ cells/ml. Cells contain 44-46% of protein and 16- 18% of lipid. The highest percent of 20:5 w3 was recorded in July (19% of total fatty acids). In the autumn, the culture density was only 200 10^ cells/ml. the protein content decreases to 33%, lipids increased to 32% and amount of 20:5w3 dropped to 3,5%. The decline of nitrogen level in the medium from 250 to 120 mg N~ ' lowered the protein level by 6.5%. The highest amount of protein was found in cells which were grown using the combination of ammoniacal and urea nitrogen. The amount of lipids can reach 40-45% under unfavourable conditions (e.g. decreasing tempera- ture, low level of both irradiance and nitrogen). EFFECTS OF THE TOXIC DINOFLAGELLATE, PFIESTE- RIA PISCICIDA, ON JUVENILE BAY SCALLOPS {AR- GOPECTEN IRRADIANS, LAMARCK). Jeffrey J. Springer* and JoAnn Burkholder, Department of Botany, North Carolina State University. Raleigh. NC 27695; Sandra E. Shumway, Di- vision of Natural Sciences. Southampton College of Long Island University, Southampton. NY 11968. The recently discovered toxic dinoflagellate. Pfiesieria pisci- cida. has been implicated as the causative agent in at least 50% of the major finfish kills (> 1,000 fish) since 1991 along the North Carolina coast. Preliminary data indicate that a neurotoxin re- leased by this dinoflagellate adversely affects certain species of shellfish as well. Zoospores oi P. piscicida release a yet-to-be characterized neu- rotoxin that may render a fish helpless in minutes. Sloughed scales and tissue from the dying fish are then fed upon by the zoospores. In the period between a fish kill. Pfiestenu piscicida may encyst aiid remain in the sediments until the next bloom is triggered. It can also survive in a multitude of forms including amoebae as well as non-toxic zoospores (NZs), subsisting heterotrophically on diets of flagellated algal prey. However, those forms which do not encyst can transform into toxic zoospores (TZs) within minutes after being introduced to fish. Release of the toxin(s) associated with P. piscicida is known to be lethal to finfish both in the field and laboratory when di- noflagellate cells are present in sufficient densities (>300 cells/ ml). This toxin has also been documented to cause adverse neu- rological and immunological effects in humans. While a major fish kill event involving P. piscicida is occur- ring, floating fish carcasses are highly visible and tend to draw public interest. However, what is not known is the effects of the dinoflagellates zoospores on shellfish located in the general vi- cinity of the kill. The goal of this research was to determine short-term effects of zoospores on shellfish popultations in estua- rine ecosystems. A grazing study was conducted to assess the potential impact on commercially important shellfish as indicated by the bay scallop. Argopecten irradians. A significant decrease in clearance rate was noted along with a '"narcotizing" effect on exposed scallops. Some scallops ceased feeding after 15 minutes of continuous exposure to zoospores. The effects of long-term exposure to zoospores must be studied in further research to de- termine adverse impacts on shellfish populations in areas repeat- edly affected by toxic outbreaks off. piscicida. DEVELOPMENTAL SHIFTS IN THE FEEDING BIODY- NAMICS OF JUVENILE UTTERBACKIA IMBECILIS (MOLLUSCA: BIVALVIA). R. A. Tankersley.* Department of Biological Sciences. University of Maryland. Baltimore County. MD; J. J. Hart and M. G. Wieber. Biology Depart- ment. Gonzaga University. Spokane. WA. Ontogenetic shifts in the feeding mechanisms utilized by juve- nile mussels iUllcrlnickici inihecilis) immediately following trans- formation were determined and associated with morphological changes in pallial feeding structures. Video recordings of feeding activities indicated juvenile U. imbecilis utilize a combination of interstitial suspension and deposit feeding to capture and ingest particles. Cilia located on the foot, gills, and anterior edge of the mantle produce anterior inhalant currents that draw suspended par- ticles into the mantle cavity for ingestion. Deposited particles were collected and drawn toward the pedal gape using both pedal-sweep and pedal locomotory feeding. The relative contribution of each feeding mode to the ingestion rate of 8. 14 and 24 day old juve- niles was deteniiined by examining the gut contents of mussels fed fluorecently labeled latex beads. Dominant feeding mode varied with age. with younger juveniles relying more heavily upon de- posit feeding mechanisms than older mussels. The rate of deposit feeding was enhanced by the presence of fine silt (<202 (xm), suggesting that particles too large to ingest may serve as important substrata for deposit feeding. Ontogenetic shifts in the mode of particle acquisition were accompanied by changes in the functional National Shellfisheries Association. Baltimore, Maryland Abstract. 1996 Annual Meeting. April 14-18. 1996 531 morphology of suspension feeding structures, including the and number of ctenidial filaments and ciliary tracts. size WHAT A YEAR TO BE A MUD CRAB! THREE YEARS ON THE BAY SCALLOP RESTORATION PROJECT, WEST- PORT RIVER ESTUARY, MASSACHUSETTS. Wayne H. Turner.* Karin A. Tammi, and Bethany A. Starr. The Water Works Group, Inc., Post Office Box 197, Westport Point, MA 02791. The Bay Scallop Restoration Project (BSRP) was launched in 1993 with the goal of generating the interest, involvement, and enthusiasm required to restore and enhance the renewable economic resources of traditional fishing and farming communi- ties. 83.000 hours of volunteer work enthusiastically invested by teams of people have been channeled into this effort. These peo- ple: students, teachers, parents, graduate students — have left a major impact, not only on the bay scallop. Argopecten irradians. but on the way in which communities participate and positively affect the direction of their economic future and environmental quality. With three years of research on the BSRP. community volun- teers, led by graduate students have uncovered several significant clues about bay scallop propagation in the Westport River. In 1993. when the BSRP first began, mud crabs, Panopeus spp.. went largely unnoticed as very few were found on the river bottom or in the propagation equipment. Green crabs. Carcinus maenas. on the other hand, were plentiful and practically every spat bag (propagation equipment used to catch juvenile scallops) had at least one if not two green crabs associated with it. Strangely enough, in the summer of 1994. green crabs took a dive and mud crabs surged. Researchers began counting thousands of mud crabs as they poured from nearly every spat bag. Because of the recent prevalence of mud crabs, an experiment using float- ing rafts was set up in the Westport River; each raft housing a different combination of four ingredients: mud crabs, green crabs. spat size bay scallops, and yearling tautog. Tautoga onitis (ap- proximately two inches in length). The conclusions drawn from this study are intriguing: 1 ) mud crabs ate the bay scallops; 2) green crabs did not eat the bay scallops and instead ate the mud crabs; 3) yearling tautog could not seem to handle a green crab (probably due to size differences). but cleverly enough, researchers observed that mud crabs in a raft with yearling tautog lost one leg per day. By the fourth day of the experiment, the mud crab could no longer move and the tautog ate it. 1995, therefore appears to be a good year to be a mud crab for several reasons. First, green crabs have been down in numbers for the past two years. Secondly, yearling tautog, com- monly found in spat bags in 1993 were virtually absent during 1994 and 1995. Finally, supporting this abundant supply of mud crabs is the propagation activities of the BSRP which is increas- ing bay scallop spat production, a preferable food source for mud crabs. FECUNDITY ESTIMATES OF THE SOUTHERN SURF- CLAM SPISULA SOLIDISSIMA SIMILIS. Randal L. Walker,* Dorset H. Hurley, and Michelle L. Jansen, Shellfish Aquaculture Laboratory. University of Georgia Marine Extension Service. 20 Ocean Science Circle, Savannah, GA 31411-1011. Fecundity estimates for two stocks of Spisula solidissima si- itiilis (Say. 1822) were determined in laboratory spawning trials. One stock was dredged from St. Catherines Sound, but trans- planted to field grow-out cages planted in a sand flat at the mouth of House Creek, Little Tybee Island, Wassaw Sound, for six months prior to the spawning trial. A second stock was dredged from St. Catherines Sound prior to the spawning trial. One-year- old clams were injected weekly with 0.2 ml serotonin in the pos- terior adductor muscle to induce spawning from March 29, 1995 until the end of June 1995. For clams ranging in shell length from 26 to 50 mm. egg production per female ranged from 0.14 to 13 million eggs. The House Creek stock produced a greater mean number of eggs per female (4.8 x 10*" versus 1.7 x 10*^) and spawned 2.8 times more than St. Catherine's Sound stock. No relationship of number of eggs per female to shell length occurred for either stock. Overall, eggs from the House Creek stock (x = 59. 1 |xm) were significantly larger (p < 0.0(X)1 ) than eggs (x = 57.4 jim) from the St. Cathenne's Sound stock. Although small in size, one- year-old southem surfclams can produce sufficient numbers of eggs per female to be utilized as brood stock for the development of an aquaculture fishery for this subspecies in the southeastern U.S. PREPARED FOOD COUPLED WITH MANIPULATION OF PHOTOPERIOD YIELD AN OUT-OF-SEASON CROP FOR THE NORTHEASTERN SEA URCHIN. Charles W. Walker and Michael P. Lesser, Department of Zoology. University of New Hampshire, Durham, NH 03824. The fishery for the green sea urchin (Strongylocentrotus droe- bachiensis) has rapidly grown to become the second largest in the Northeastern United States behind lobsters. Overfishing has dras- tically depleted once abundant natural populations. Two other problems naturally plague the industry. One of these is poor roe quality in a large percentage of the urchins harvested, leading to a lower than maximum price. Another is the short period when roe quality is high. There is a window of time from September until February when urchins have firm, ripe gonads. If urchins in a land based aquaculture facility could be fed a prepared food and be induced to ripen again after February, then the period of availabil- ity of highest quality roe could be expanded, greatly increasing the market potential for Gulf of Maine urchin roe. We have coupled: 1 ) enhancement of gonadal growth of poorly fed urchins utilizing prepared food with 2) photoperiodic manipulation of the gameto- genic cycle to produce an out-of-season crop which could be used to exploit a lucrative end of summer market now supplied by Chile. Rather spawned urchins (March. 1995; =£6% gonad index) were held under artificial illumination, using astronomic clocks set to simulate June photoperiod and were feed 3 g prepared food/ 532 Abstract. 1996 Annual Meeting, April 14-18, 1996 National Shellfisheries Association, Baltimore, Maryland animal/week. This resulted in a significant increase in gonad size compaired with field populations (3=25%) without a corresponding increase in test size. Histological examination of monthly samples of gonads indicates that this growth is a result of increase in size of nutritive phagocytes (which are intragonadal nutrient storage cells) yielding significantly higher gonadal indices than those si- multaneously observed in field populations. After 3 months on this feeding regime, urchins were then exposed to September photo- period which is known to naturally stimulate gametogenesis for urchins in the field. Stereological analysis of histological sections, indicate that spermatogonia in such animals undergo rapid prolif- eration and normal spermatogenesis three months early. Oogonia also proliferate early, but resulting oocytes undergo minimal vi- tellogenesis. Testes and ovaries both increase in size to gonad indices of 28-30% which is based on accumulation of normal spermatozoa in males and continued growth of nutritive phago- cytes in females. Supported by Sea Grant Development Funds. New Hampshire and Hatch Grant #353 to C. W. Walker and M. P. Lesser. MYELOPEROXIDASE ACTIVITY FROM BLOOD CELLS OF THE EASTERN OYSTER. CRASSOSTREA VIRGINICA. Jennifer Wojcik* and Kennedy T. Paynter, Department of Zo- ology, University of Maryland. College Park, MD 20742. Hemocytes of most bivalve molluscs are amoeboid, phagocytic cells which comprise an important part of the bivalve immune system. Similar to vertebrate macrophages, hemocytes engulf in- vading or other non-host entities and attack them within the pha- gosome through a series of biochemical reactions including lysos- ome and protease activation, and reactive oxygen intermediate (ROD production. There are four types of ROls typically pro- duced; superoxide anions (02" ), hydrogen peroxide (H202), hy- droxyl radicals (OH), and hypochlorous acid (HOCI). In oysters, oxidative killing is thought to be one of the primary forms of immune response. In most phagocytic cells, myeloperoxidase (MPO) catalyzes the production of hypochlorous acid from hydrogen peroxide and chloride. Hemocytes from a number of molluscan species have been shown to produce reactive oxygen metabolites, which could serve as a substrate for myeloperoxidase. When cells engulf the foreign particles, enzymes are brought into the phagosome and are then activated as the pH decreases. The enzymes catalyze the production of ROls which have damaging biochemical effects on the foreign cells. ROl production is quantified by using a chemilu- minescent assay. Taurine, a scavenger of hypochlorous acid, com- pletely quenched chemiluminescence in oyster hemocytes, indi- cating that the cells produce hypochlorous acid during and shortly after phagocytosis. We measured myeloperoxidase activity in extracts of oyster hemocytes using a variety of techniques. Tetramethylbenzidine (TMB) was peroxidized readily by small amounts of hemocyte extract, indicating a significant amount of MPO may be present. TMB peroxidation had a pH optimum of approximately 5.5. At- tempts to differentiate between halide-independent and halide- dependent activities using diethanolamine and a taurochloramine assay yielded ambiguous results. Other techniques are currently being used to assess putative MPO activity and its importance in the oysters cytotoxic response. THE NATIONAL SHELLFISHERIES ASSOCIATION The National Shellfisheries Association (NSA) is an international organization of scientists, manage- ment officials and members of industry that is deeply concerned and dedicated to the formulation of ideas and promotion of knowledge pertinent to the biology, ecology, production, economics and man- agement of shellfish resources. The Association has a membership of more than 1000 from all parts of the USA, Canada and 18 other nations; the Association strongly encourages graduate students' mem- bership and participation. WHAT DOES IT DO? — Sponsors an annual scientific conference. — Publishes the peer-reviewed Journal of Shellfish Research. — Produces a Quarterly Newsletter. — Interacts with other associations and industry. WHAT CAN IT DO FOR YOU? — You will meet kindred scientists, managers and industry officials at annual meetings. — You will get peer review through presentation of papers at the annual meeting. — If you are young, you will benefit from the experience of your elders. — If you are an elder, you will be rejuvenated by the fresh ideas of youth. — If you are a student, you will make useful contacts for your job search. — If you are a potential employer, you will meet promising young people. — You will receive a scientific journal containing important research articles. — You will receive a Quarterly Newsletter providing information on the Association and its activities, a book review section, information on other societies and their meetings, a job placement section, etc. HOW TO JOIN — Fill out and mail a copy of the application blank below. The dues are 45 US $ per year ($25 for students) and that includes the Journal and the Newsletter! NATIONAL SHELLFISHERIES ASSOCIATION^APPLICATION FOR MEMBERSHIP (NEW MEMBERS ONLY) Name: For the calendar year: Date: Mailing address: Institutional affiliation, if any: Shellfishery interests: Regular or student membership: Student members only — advisor's signature REQUIRED: Make cheques (MUST be drawn on a US bank), international postal money orders or VISA for $45 ($25 for students with advisor's signature) payable to the National Shellfisheries Association and send to Christine Hodgson, BC Ministry of Ag. and Fisheries, 2500 Cliffe Avenue, Courtenay, British Columbia, CANADA V9R 5M6. INFORMATION FOR CONTRIBUTORS TO THE JOURNAL OF SHELLFISH RESEARCH Original papers dealing with all aspects of shellfish re- search will be considered for publication. Manuscripts will be judged by the editors or other competent reviewers, or both, on the basis of originality, content, merit, clarity of presentation, and interpretations. Each paper should be carefully prepared in the style followed in Volume 13. (1). of the Journal of SheUjish Research ( 1991 ) before sub- mission to the Editor. Papers published or to be published in other journals are not acceptable. Title, Short Title. Key Words, and Abstract: The title of the paper should be kept as short as possible. Please include a "'short running title" of not more than 48 char- acters including space between words, and approximately seven (7) key words or less. Each manuscript must be ac- companied by a concise, informative abstract, giving the main results of the research reported. The abstract will be published at the beginning of the paper. No separate sum- mary should be included. Text: Manuscripts must be typed double-spaced throughout on one side of the paper, leaving ample margins, with the pages numbered consecutively. Scientific names of species should be underlined and. when first mentioned in the text, should be followed by the authority. Common and scientific names of organisms should be in accordance with American Fisheries Society Special Publications 16 and 17; Common and Scientific Names of Aquatic Invertebrates from the United States and Canada: MoUusks and CSNAIUSC: Decapod Crustaceans. Abbreviations, Style, Numbers: Authors should follow the style recommended by the fourth edition (1978) of the Council of Biology Editors [CBE] Style Manual, dis- tributed by the American Institute of Biological Sciences. All linear measurements, weights, and volumes should be given in metric units. Tables: Tables, numbered in Arabic, should be on sepa- rate pages with a concise title at the top. Illustrations: Line drawings should be in black ink and planned so that important details will be clear after reduc- tion to page size or less. No drawing should be so large that it must be reduced to less than one third of its original size. Photographs and line drawings preferably should be pre- pared so they can be reduced to a size no greater than 17.3 cm X 22.7 cm, and should be planned either to occupy the full width of 17.3 cm or the width of one column, 8.4 cm. Photographs should be glossy with good contrast and should be prepared so they can be reproduced without re- duction. Originals of graphic materials (i.e., line drawings) are preferred and will be returned to the author. Each illus- tration should have the author's name, short paper title, and figure number on the back. Figure legends should be typed on separate sheets and numbered in Arabic. No color illustrations will be accepted unless the author is prepared to cover the cost of associated reproduction and printing. References Cited: References should be listed alphabet- ically at the end of the paper. Abbreviations in this section should be those recommended in the American Standard for Periodical Title Abbreviations, available through the American National Standard Institute, 1430 Broadway, New York, NY 10(M8. For appropriate citation format, see examples at the end of papers in Volume 10, Number I , of the Journal of Shellfish Research or refer to Chapter 3, pages 51-60 of the CBE Style Manual. Page Charges: Authors or their institutions will be charged $65.00 per printed page. If illustrations and/or tables make up more than one third of the total number of pages, there will be a charge of $30.00 for each page of this material (calculated on the actual amount of page space taken up), regardless of the total length of the article. All page charges are subject to change without notice. Proofs: Page proofs are sent to the corresponding author and must be corrected and returned within seven days. Al- terations other than corrections of printer's errors may be charged to the author(s). Reprints: Reprints of published papers are available at cost to the authors. Information regarding ordering reprints will be available from The Sheridan Press at the time of printing. Cover Photographs: Particularly appropriate photo- graphs may be submitted for consideration for use on the cover of the Journal of Shellfish Research. Black and white photographs, if utilized, are printed at no cost. Color illus- trations may also be considered. Corresponding: An original and two copies of each manuscript submitted for publication consideration should be sent to the Editor, Dr. Sandra E. Shumway, Natural Science Division. Southampton College, LIU Southamp- ton, NY 11968, Ph. 516-287-8407, FAX 516-287-8419. email: sshumway(S'sunbum. liunet.edu Membership information may be obtained from the Trea- surer using the form in the Journal. Institutional subscribers should send requests to: Journal of Shellfish Research, P.O. Box 465. Hanover, PA 17331. Shelley A. Burton, Gerald R. Johnson and T. Jeffrey Davidson Cytologic sexing of marine mussels (Mylilus edulis) 345 M. Jose Fernandez-Reiriz, Uxio Labarta and Jose M. F. Babarro Comparative allometries in growth and chemical composition of mussel Mytilus galloprovincialis, (Lamarck) cultured in two zones in the Ria Sada (Galicia, NW Spain) 349 Masahiro Nakaoka Size-dependent survivorship of the bivalve Yoldia notabilis (Yokoyama, 1920); the effect of crab predation 355 A. Figueras, J. A. F. Robledo and B. Novoa Brown ring disease and parasites in clams (Rudiiapes decussatus and R. philippinarum) from Spain and Portugal 363 C. Riquelme, G. Hayashida, R. Araya, A. Uchida, M. Satomi and Y. Ishida Isolation of a native bacterial strain from the scallop Argopecten purpuraius with inhibitory effects against pathogenic vibrios 369 Fu-Lin E. Chu, Aswani K. Volety and Gegorgeta Constantin A comparison of Crassostrea gigas and C. virginica: effects of temperature and salinity on susceptibility to the protozoan parasite, Perkinsus marinus 375 Gustavo W. Calvo, Robbie J. Fagan, Kelly N. Greenhawk, Gary F. Smith and Stephen J. Jordan Spatial distribution and intensity of Perkinsus marinus infections in oyster recovery areas in Maryland 381 Thomas C. Cheng and John J. Manzi Correlation between the presence of lathyrose with the absence of Haplosporidium nelsoni in Crassostrea virginica from two South Carolina tributaries where Perkinsus marinus also inhibits hemocyte agglutination by the Lathyrus odoratus lectin 391 Adriana I. Zabaleta and Bruce J. Barber Prevalence, intensity, and detection of Bonamia ostreae in Ostrea edulis L. in the Damariscotta River area, Maine ... 395 Islay D. Marsden and Paul M. J. Williams Factors affecting the grazing rate of the New Zealand abalone Haliotis iris Martyn 401 Allan W. Stoner, Robert A. Glazer and Peter J. Barile Larval supply to queen conch nurseries; relationships with recruitment process and population size in Florida and the Bahamas 407 Volker Koch and Matthias Wolff The mangrove snail Thais kiosquiformis Duclos; a case of life history adaptation to an extreme environment 42 1 P. B. Brown, M. R. White, J. Chaille, M. Russell and C. Oseto Evaluation of three anesthetic agents for crayfish (Orconectes virilis) 433 Jean-Marie Sevigny and Bernard Sainte-Marie Electrophoretic data support the last-male sperm precedence hypothesis in the snow crab, Chionoecetes opilio (Brachyura; Majidae) 437 J. Stephen Hopkins, Paul A. Sandifer, Craig L. Browdy and John D. Holloway Comparison of exchange and no-exchange water management strategies for the intensive pond culture of marine shrimp 44 1 Abstracts of technical papers presented at the 16th Annual Aquaculture Seminar, Milford, Connecticut, February 26-28, 1996 447 Abstracts of technical papers presented at the 88th Annual Meeting of the National Shellfisheries Association, April 14-18, 1996, Baltimore, Maryland 465 COVER PHOTO: Thurlow Nelson on board the Julius Nelson ca. 1948-1949. Photo courtesy of Haskin Shellfish Research Laboratory. The Journal of Shellfish Research is indexed in the following; Science Citation Index®, Sci Search®, Research Alert®, Current Contents®/ Agriculture, Biology and Environmental Sciences, Biological Abstracts. Chemical Abstracts, Nutrition Abstracts, Current Advances in Ecological Sciences, Deep Sea Research and Oceanographic Literature Review, Environmental Periodicals Bibliography, Aquatic Sciences and Fisheries Abstracts, and Oceanic Abstracts. JOURNAL OF SHELLFISH RESEARCH Vol. 15, No. 2 JUNE 1996 CONTENTS IN MEMORIAM R. Tucker Abbott (1919-1995) 1 85 C. E. Carver, A. L. Ballet, R. Warnock and D. Douglas Red-coloured digestive glands in cultured mussels and scallops: the implication oi Mesodinium riibrum 191 Jan H. iMndsberg Neoplasia and biotoxins in bivalves: Is there a connection? 203 John C. Wekell, Roseanne M. Lorenzana, Mara Hogan and Harold Barnett Survey of paralytic shellfish poison and domoic acid in Puget Sound predatory gastropods 23 1 Wayne A. O'Connor and Michael P. Heasman Temporal patterns of reproductive condition in the doughboy scallop, Chlamys (Mimachlamys) asperrima Lamarck, in Jervis Bay, Australia 237 G. Martinez, C. Garrote, L. Mettifogo, H. Perez and E. Uribe Monoamines and prostaglandin E, as inducers of the spawning of the scallop, Argopecten purpuratus Lamarck 245 Joan L. Manuel, Susan Burbridge, Ellen L. Kenchington, Martin Ball and Ronald K. O'Dor Veligers from two populations of scallop (Placopecten magellanicus) exhibit different vertical distributions in the same mesocosm 25 1 Peter Coutteau, John D. Castell, Robert G. Ackman and Patrick Sorgeloos The use of lipid emulsions as carriers for essential fatty acids in bivalves: a test case with juvenile Placopecten magellanicus 259 Mohsin U. Patwary, Michael Reith and Ellen L. Kenchington Isolation and characterization of cDNA encoding an actin gene from sea scallop (Placopecten magellanicus) 265 Robert B. Rheault and Michael A. Rice Food-limited growth and condition index in the eastern oyster, Crassostrea virginica (Gmelin, 1791), and the bay scallop, Argopecten irradians irradians (Lamarck, 1819) 271 Bruce J. Barber Gametogenesis of eastern oysters, Crassostrea virginica (Gmelin, 1791) and Pacific oysters, Crassostrea gigas (Thunberg, 1793) in disease endemic lower Chesapeake Bay 285 Gregory A. DeBrosse and Standish K. Allen, Jr. The suitability of land-based evaluations of Crassostrea gigas (Thunberg, 1793) as an indicator of performance in the field 29 1 Federico Garcia-Dominguez, Bertha Patricia Ceballos-Vasquez and Arturo Tripp Quezada Spawning cycle of the pearl oyster, Pinctada mazatlanica (Hanley, 1856), (Pteriidae) at Isla Espiritu Santo, Baja California Sur, Mexico 297 A. G. Jeffs and R. G. Creese Overview and bibliography of research on the Chilean oyster Tiostrea chilensis (Philippi, 1845) from New Zealand waters 305 Francis X. O'Beirn, Randal L. Walker and Peter B. Heffernan Enhancement of subtidal eastern oyster, Crassostrea virginica (Gmelin, 1791), recruitment using mesh bag enclosures 313 Mijin Lee, Gordon T. Taylor, V. Monica Bricelj, Susan E. Ford and Steve Zahn Evaluation of Vibrio spp. and microplankton blooms as causative agents of juvenile oyster disease in Crassostrea virginica (Gmelin, 1791) 319 Malcolm Haddon, Trevor J. Willis, Robert G. Wear and Victor C. Anderlini Biomass and distribution of five species of surfclam off an exposed west coast North Island beach. New Zealand 331 John W. Crenshaw, Jr., Peter B. Heffernan and Randal L. Walker Effect of growout density on heritability of growth rate in the northern quahog, Mercenaria mercenaria (Linnaeus, 1758) 341 CONTENTS CONTINUED ON INSIDE BACK COVER JOURNAL OF SHELLFISH RESEARCH VOLUME 15, NUMBERS DECEMBER 1996 The Journal of Shellfish Research (formerly Proceedings of the National Shellfisheries Association) is the official publication of the National Shellfisheries Association Editor Dr. Sandra E. Shumway Natural Science Division Southampton College, LIU Southampton, NY 11968 Dr. Standish K. Allen, Jr. (1998) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Peter Beninger (1997) Department of Biology University of Moncton Moncton, New Brunswick Canada El A 3E9 Dr. Andrew Boghen (1997) Department of Biology University of Moncton Moncton. New Brunswick Canada ElA 3E9 Dr. Neil Bourne (1996) Fisheries and Oceans Pacific Biological Station Nanaimo, British Columbia Canada V9R 5K6 Dr. Andrew Brand (1996) University of Liverpool Marine Biological Station Port Erin, Isle of Man Dr. Eugene Burreson (1997) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Peter Cook (1998) Department of Zoology University of Cape Town Rondebosch 7700 Cape Town, South Africa EDITORIAL BOARD Dr. Simon Cragg (1998) Faculty of Technology Buckinghamshire College of Higher Education Queen Alexandra Road High Wycombe Buckinghamshire HP 11 2JZ England, United Kingdom Dr. Leroy Creswell (1997) Harbor Branch Oceanographic Institute US Highway 1 North Fort Pierce, Florida 34946 Dr. Lou D'Abramo (1998) Mississippi State University Dept of Wildlife and Fisheries Box 9690 Mississippi State, Mississippi 39762 Dr. Ralph Elston (1996) Battelle Northwest Marine Sciences Laboratory 439 West Sequim Bay Road Sequim, Washington 98382 Dr. Susan Ford (1998) Rutgers University Haskin Laboratory for Shellfish Research P.O. Box 687 Port Norris, New Jersey 08349 Dr. Raymond Grizzle (1997) Randall Environmental Studies Center Taylor University Upland, Indiana 46989 Dr. Robert E. Hillman (1998) Battelle Ocean Sciences New England Marine Research Laboratory Duxbury, Massachusetts 02332 Dr. Mark Luckenbach ( 1997) Virginia Institute of Marine Science Wachapreague, Virginia 23480 Dr. Bruce MacDonald (1997) Department of Biology University of New Brunswick P.O. Box 5050 Saint John, New Brunswick Canada E2L 4L5 Dr. Roger Mann (1998) Virginia Institute of Marine Science Gloucester Point, Virginia 23062 Dr. Islay D. Marsden (1996) Department of Zoology Canterbury University Christchurch, New Zealand Dr. Kennedy Paynter (1998) 1200 Zoology Psychology Building College Park, Maryland 20742-4415 Dr. Michael A. Rice (1996) Dept. of Fisheries, Animal & Veterinary Science The University of Rhode Island Kingston, Rhode Island 02881 Dr. Tom Soniat (1998) Biology Department Nicholls State University Thibodaux, Louisiana 70310 Susan Waddy (1997) Biological Station St. Andrews, New Brunswick Canada, EOG 2XO Mr. Gary Wikfors (1998) NOAA/NMFS Rogers Avenue Milford, Connecticut 06460 Journal of Shellfish Research Volume 15, Number 3 ISSN: 00775711 December 1996 ., i \03l Journal of Shellfish Research. Vol. 15. No. 3. 533-534. 1996. Robert R. L. Guillard Honored Life Member National Shellfisheries Association As we honor Dr. Robert R. L. Guillard — Bob — with lifetime membership in the National Shellfisheries Association, we should remind ourselves of two important aspects of Bob's relationship with shellfish. The first is that the rearing ot shellfish in captivity, both for expenmental research and aquaculture production, would not be possible without Bob's pioneering work in phytoplankton culture. The second important point is that Bob's contributions to the world of shellfish, although not exactly inadvertent, are no more than fortuitous offshoots of his research focus on the physiological ecology of phytoplankton. We should all hope that our sidelines are so successful! Bob began his professional life as an electrical engineer at the Navy Yard in New York City; perhaps mussels set on the hulls, but otherwise, this seems a long way from shellfish. Graduate studies in microbial ecology at Yale, leading to a Ph.D. in 1954, brought Bob to Connecticut, where he became acquainted with Victor Loosanoff. Dr. Loosanoff then was Director of the U.S. Fish and Wildlife Service's Milford Marine Biological Laboratory and a fixture at Yale marine science seminars, having completed his own Ph.D. there. Apparently, Loosanoff would preface all questions of seminar speakers in his strong Russian accent, "As you know, I am interested from oysters ..." Efforts to grow oyster larvae at the Milford Laboratory on fertilized, bloomed seawater had met with limited success. and communication with the Plymouth Laboratory suggested advantages of feeding selected phytoplankton to larval shellfish. Loosanoff must have seen in one student. Bob Guillard. the expertise needed to produce baby food for his oysters: a position funded by the Oyster Institute of North America was secured for Bob to spend several years at Milford. During his time at Milford. Bob isolated a number of the phytoplankton cultures used widely to this day in marine research and shellfish culture, including 3H Thalassiosira pseudonana . and Synechococcus baciUans (a cyanobacterium that would revolutionize biological oceanography 20 years later). Bob Guillard's first full research report was an article published in 1957, not coincidentally in the Proceedings of the National Shellfisheries Associalion (Vol. 48, pp. 134-142), titled, "Some Factors in the Use of Nannoplankton Cultures as Food for Larval and Juvenile Bivalves." Between this article and the 1958 USFWS Fish. Bull. 136 (Vol. 58), "Relative Value of Ten Genera of Microorganisms as Foods for Oyster and Clam Lar\'ae," by Harry Davis and Bob Guillard. most of the practical information we use to this day in deciding what phytoplankton to feed molluscan larvae was established. Countless studies of basic shellfish biology, not to mention the establishment of hatchery-based shellfish aquaculture. were made possible by Bob Guillard's identification of practical algal diets. If Bob Guillard did nothing more to benefit the shellfish community, his place among the legendary figures of shellfish biology would be assured. Then, he invented f/2. In July 1958, Bob had accepted a research position at Woods Hole Oceanographic Institution. While working to establish a collection of marine phytoplankton cultures for studies of plankton ecology. Bob faced the challenge of developing a nutrient enrichment for seawater that would support survival and growth of the widest possible range of microalgal taxa — no mean feat, considering the physiological diversity represented. Achievement of the "right recipe" was coincident with completion of a study. 533 534 Honored Life Member: Robert R. L. Guillard published with John Ryther in 1962. having the seemingly arcane title, "Studies of Marine Planktonic Diatoms. I. Cyclotella nana Hustedt and Detonula cunfervaceae (Cleve) Gran." in the Canadian Journal of Microbiology (Vol. 8, 229-239). This article would become one of the most-cited in marine science, not because of extreme interest in the two diatoms that dominate the title, but because the seawater enrichment detailed in this report — designated f/2 — turned out to be the most successful algal-culture medium ever developed, "f/2" has trademark recognition in marine science that would be the envy of most breakfast cereals, and a number of aquaculture-supply companies market premixed products of this composition. For this contribution. Bob Guillard does not deserve to be merely famous (which he is anyway), but he deserves to be very wealthy! A move to the Bigelow Laboratory for Ocean Sciences in West Boothbay Harbor, ME, in 1982 led to the establishment of the Provasoli-Guillard Center for the Culture of Marine Phytoplankton (CCMP) there in 1985. Bob was director of this institution from its mception until his "retirement" in 1989. CCMP brought together the two great, privately held U.S. collections of marine phytoplank- ton. Bob's Woods Hole Collection and that of the late Dr. Luigi Provasoli of the famed Haskins Laboratories. The Center also established a framework and organization to ensure that phytoplankton strains of known origin and identity are available to the research community far into the future. Again, we have Bob Guillard to thank for building an algal supermarket where we can shop for shellfish munchies and know we are getting what we ask for. As a publication record of well over 100 articles and many book chapters, notes, abstracts, etc. attests. Bob Guillard continues to contribute mightily to the field of phytoplankton ecology. In his "retirement," Bob has only increased the pace of his scientific activities, having shaken loose the chains of bureaucracy that inevitably accompany the title, "Director." Bob continues, as he has for many years, to teach. There are the legendary short courses. There are the endless telephone calls for help to which we subject him; there is no known instance of anyone whose call for help was ignored or given short shrift. There are the endless telephone calls he makes to students and professionals alike to suggest research directions and ideas — these seem to pop into his head much faster than even he can follow up on. There is the boundless cunosity and enthusiasm that continue to inspire. A sign of maturity is the ability to articulate ideas in direct, simple language. Bob expresses the question driving his work with phytoplankton as, "Why do they live where they do?" This seemingly simple question weaves physical, chemical, and biological threads into the fabric of the invisible ecosystem — that of organisms too small for us to see, catch, dissect, and catalog with our unaided senses. To cultivate, using the limited resolution of the microscope and a dizzying array of indirect methods of measurement, a garden in which the smallest flowers will thrive requires a rare combination of knowledge, insight, and intuition. Bob Guillard has brought these talents to bear on the challenges of culturing phytoplankton . . . relentlessly. As we honor him at our annual meeting. Bob is back in his laboratory nursing along another new "bug," one that may reveal more secrets of his invisible ecosystem. We wish him luck with it and, for our own sakes. hope that it is the perfect food for larval oysters. Bob has shared with some of us the irony of his interactions with shellfish farmers. He describes the typical telephone troubleshooting scenario as a series of phone calls in which the hatchery operator relates a problem and Bob suggests a response. The hatchery operator calls back to say that the suggested action did not fix the problem. Bob suggests the next step. The process is repeated. "Eventually," says Bob, "they stop calling. That's how 1 know what finally worked." Bob Guillard, as we thank you for over 40 years of help fixing our most difficult problems, we want you to know. "IT WORKED!" Gary H. Wikfors Milford. CT Jminml of ShcUfish Research. Vol. 15. No. 3, 535-541, IW6. EFFECTS OF DIFFERENT SUBSTRATA AND PROTECTIVE MESH BAGS ON COLLECTION OF SPAT OF THE PEARL OYSTERS, PINCTADA MARGARITIFERA (LINNAEUS, 1758) AND PIN CT ADA MACULATA (GOULD, 1850)' KIM J. FRIEDMAN AND JOHANN D. BELL International Centre for Livini^ Aquatic Resource Management (ICLARM) Coastal Aquacullure Centre P.O. Box 43H Honiara. Solomon Islands ABSTR.-iCT Refining techniques for the collection of spat is important to the culture of blacklip pearl oysters, Pinclada margari- tiferu. especially where the collection of spat is marginally effective. We deployed 40 spat collectors at 15 sites within the open reef complexes of Solomon Islands to test the effects of different collectors (constructed of shademesh and plastic sheeting) and protective mesh bags on the abundance of spat. After 6 mo, we recorded abundances of P. margariiifera. and another pearl oyster. P. maculaiu. together with the numbers of predators associated with the collectors. Significantly more P. mcirftaritlfera were found on the shademesh. whereas live P. maciilala were more abundant on the plastic sheeting. Collectors inside protective mesh bags did not yield more pearl oysters than those left unprotected. Mesh bags trapped predators such as Cynuitiiim spp. gastropods and portunid crabs settling to the collectors from the plankton. The bags also fouled easily, impeding waterflow to the collector. We conclude that experiments should be conducted to identity optimal materials for collecting the target species of pearl oyster and that collectors should not be placed in protective mesh bags in environments similar lo those of Solomon Islands. KEY WORDS: Aquaculturc. pearl oysters. Pinclada. spat, settlement, substrates 'ICLARM contribution no. 1240. INTRODUCTION In the past 20 y, there has been rapid expansion in the farming of blacklip pearl oysters. Pinclcuhi mcirt^antifeni. for the culture of black pearls (Intcs and Coeroli 1985). The expansion of this in- dustry has been particularly pronounced in French Polynesia and Cook Islands (Rowntree 1993), where it rivals tourism as the major source of foreign exchange. The culture of blacklip pearl oysters in French Polynesia and Cook Islands was based initially on the use of wild shell from the lagoons of selected atolls: col- lection of spat provided only a minor proportion of the farmed shell (Coeroli et al. 1984). In the last 1980s and early 1990s, however, legislation was introduced to parts of French Polynesia and Cook Islands banning the use of wild shells. Consequently. the industry became more dependent on the collection of spat to provide the oysters needed for pearl culture. The spat of the blacklip pearl oysters are collected on subsur- face longlines. using a variety of settlement materials, ranging from branches of selected trees (Coeroli et al.. 1984. Victor 1987, Passfield 1989) to a variety of plastic sheets, ropes, and meshes (Coeroli et al. 1984, Cabral et al. 1985). The use of plastic sub- strata is now widespread because of the ease of use and durability (N. Sims. pers. comm). Spat collectors are hung at depths of 2-4 m. where settlement is greatest (Shirai 1970. Cabral et al. 1985, Sims 1993). Collectors are buoyed clear of the substrate to isolate them from benthic predators (Swift 1985). and in some cases, mesh bags are used to protect spat on the collectors from predators (Coeroli et al. 1984. Gervis and Sims 1992). In the course of a large-scale sampling program to identify spatial variation in an abundance of spat P. margaritifera in Sol- omon Islands, we designed experiments to answer two questions aimed at refining methods for the collection of pearl oyster spat. These questions were; (I) Do mesh coverings ("spat bags") in- crease the number of spat harvested from collectors? (2) Is there a difference in the number of spat harvested from collectors made of plastic sheeting and those made from shademesh? We found that the use of spat bags did not increase the number of P. margaritijera spat on collectors and that more spat were collected from shademesh than from plastic sheeting. During the experiments, large numbers of another pearl oyster. Pinctada mac- ulata. also settled on the collectors. This species, which produces baroque pearls of smaller size and value than those found in P. mari>iiriiifeni (Sims 1988). also provided a useful test for the effect of spat bags. At two of the three sites where this species settled in abundance, there were significantly fewer spat on col- lectors within the bags. METHODS Sampling Sites We deployed spat collectors at three sites in each of the five regions (i.e., a total of 15 sites) in the Solomon Islands (Fig. I). These sites encompassed the range of habitats thought to be suit- able for the settlement of P. inarf>aritifera. Sites were selected on the basis of maps and aerial photographs, on-site inspections, and information on past harvests of P. margaritijera supplied by local communities. Within sites, spat collectors were positioned where larvae were likely to be entrained by nearby channel flows or in areas that were semienclosed. Where possible, we placed collectors on the lee side of bays, where prevailing winds were likely to concentrate spat in the surface waters (Sims 1989). Longlines At each site, spat collectors were suspended from a single longline consisting of a 100-m headline (I2-mm polypropylene rope) supported every 20 m by a 30-cm-diamcter buoy. Using SCUBA, we submerged the longline to a depth of 3 m by pulling down on the attached buoys with a dropper line (lO-rrmi polypro- pylene rope). The dropper lines were secured to large colonies of coral at depths of 8-19 m. Longlines were positioned across shallow reef areas (<20 m deep), and the ends of the headline were attached to coral heads in 535 536 Friedman and Bell 7"S 9°S Solomon Islands Mundal Region' 157°E 159°E 161°E Figure 1. Map of the Solomon Islands showing the regions where spat were collected (circles). shallow water (<5 m deep). This allowed the headline to be ten- sioned easily to that all collectors could be maintained at the re- quired depth (2^ m). Spat Collectors We selected materials for the construction of spat collectors on the basis of the previous experience of Cabral et al. (1985) in French Polynesia and Sims (1989) in Cook Islands. These mate- rials were black plastic polyethylene sheeting (50 |xm thick) and black plastic shademesh (55% shade). One type of collector was constructed of each material. We constructed the plastic sheeting collector from 16 strips, measuring 10 x 100 cm. Each strip was threaded loosely, four or five times, onto a 1 10-cm length of 3-mm polypropylene line. The ends of several strips were passed through the weave of the 3-mm line to prevent the strips from "bunching" at the base of the collector once it was fouled. We made the shademesh collector from a folded single sheet of shademesh threaded four or five times onto a 1 10-cm length of 3-mm polypropylene line. Both types of collectors had the same surface area (1.6 m"). Experimental Design Twenty replicates of each type of collector were suspended from each longline. Protective mesh bags measuring 80 x 40 cm, with a mesh size of 2 x 5mm, were used to enclose 10 replicates of each type of collector. Thus, we had 10 replicates of four treatments for each longline: 10 x plastic sheeting open, 10 x plastic sheeting bagged, 10 x shademesh open, and 10 x shade- mesh bagged. These were randomly allocated to attachment points ever 2 m along each longline. Collectors were soaked for 6 mo (between January and July 1994), after which they were removed from the water and checked for spat of pearl oysters. In addition to recording data for P. margaritifera. we counted the numbers of the closely related pearl oyster, P. maculata. Data recorded for each collector were: the number of live and dead spat for P. margarilifera and P. macu- lata, the dorsoventral measurement (DVM) (NichoUs 1931) of each individual P. margaritifera, and the DVM of up to 10 live individuals of P. maculata. All measurements were made to the nearest 1 mm. Additionally, we recorded the numbers of several of the predators of juvenile pearl oysters identified by Sims (1989), Dharmaraj et al. (1987), and Govan (1994), especially Cymatium spp. gastropods and crabs. Govan (1994) recognized four species of Cymotium as predators of juvenile bivalves in Solomon Islands. We pooled counts for these species because it was difficult to distinguish among the juveniles. Some of the predators concealed within the collectors were highly mobile, e.g.. fish from the family Bulistidae. To ensure that we sampled them effectively, we used a fine-mesh harvesting bag measuring 80 x 1 30 cm to surround the collector before it was removed from the longline. We used data from the four types of spat collectors to test the following null hypotheses for both P . margaritifera and P. mac- ulata: ( 1 ) There was no difference in the abundance of spat recorded on the collectors made of plastic sheeting and shademesh, and (2) There was no difference in the abundance of spat collected from open and bagged collectors. Analysis of Data To test the null hypotheses, we used data only from sites that had relatively high numbers of spat (Table 1 ). For P. margaritif- era, we used five sites to analyze variation in the following mea- sures: total (live and dead) abundance of spat, ratio of live to total number of spat, and DVM. We also analyzed variation in the abundance and size of predators from these five sites. In general, P . maculata settled in far greater abundance than P. margaritifera. However, three sites received the majority of P. maculata spat (Table 1 ), and so, we restricted the analysis of data for this species to those sites. We analyzed variation in the total abundance of spat due to the effect of site (random factor), collecting material (fixed factor), and bag protection ( fixed factor) for both species of pearl oysters in a balanced three-way analysis of variance (ANOVA). We also analyzed the total abundance of Cymatium spp. in the same man- ner. To investigate variation in the DVM of live spat of P. mar- garitifera and P. maculata among collectors, we used data pooled across sites in a two-way ANOVA (materials x protection). Be- fore all analyses, we tested for homogeneity of variance using Cochrans C test and transformed data to log,o(x -I- 1) when variances were nonhomogeneous. In the case of data for the abundance of live P. maculata, transformation did not result in homogeneity of variances. For this variable, data were pooled for each type of collector across the three sites and were presented graphically to describe variability. Where ANOVA indicated that there were significant differences among means, the Student Neuman-Keuls (SNK) test or Tukey HSD test for unequal sample sizes was used to identify the nature of these differences. RESULTS Variation in Collections of P. margaritifera Among Sites and Collectors A total of 154 P. margaritifera spat were collected at the 15 sites (Table 1), but distribution was highly patchy. There was a significant difference in the total abundance of spat among the five sites where the greatest number of P. margaritifera were collected Collection of Pearl Oyster Spat 537 TABLE 1. Total abundance of P. margarilifera and P. maculata spal at each of three sites in five regions." Site P. margarilifera P. maculala Region Total Alive % Alive Total Alive % Alive Site Description Gela Islands 1 17" 9 52.9 88 32 .^6,4 Embayed reef 2 6 5 83,3 133 60 45,1 Open reef system 3 M' 10 71.4 1.257-' 924 73.5 Reef system beside channel Seghe 4 0 0 0 10 6 60.0 Lagoonal reef 5 0 0 0 3 2 66.7 Lagoonal reef 6 1 0 0 82 13 15.8 Inside edge of lagoonal reef beside a channel Munda 7 0 0 0 18 11 61.1 Channel side embayment 8 1 1 100 21 10 47.6 Channel side embayment 9 1 1 100 60 26 43.3 Lagoonal reef Gizo 10 27- 11 40.7 698^ 301 43.1 Open reef beside a channel 11 T 0 0 104 19 18.3 Embayed reef 12 4r 15 36.6 1.147" 81 7.1 Embayed reef South Malaita 13 37- 14 37,8 104 32 30.8 Inside edge of lagoon beside a channel 14 6 4 66,7 17 4 23.5 Lagoonal reef 15 0 0 0 1 0 0 Bay affected by large mangrove system Total 154 70 3.743 1.521 ' Indicates those sites where data were used in ANOVA. (Table 2; p = 0.006). but the SNK test could not differentiate among the means. Collections at the best site (Site 12) averaged just over one spat per collector (Table 1 ). There was a significant difference in the total abundance off. margarilifera on collectors made of shademesh and plastic sheet- ing (Table 2; p = 0.01). A mean of 0.98 (±0.13 SE) P. marga- rilifera was found on collectors made from shademesh, whereas a mean of 0.38 (± 0.06 SE) spat was collected from those made of plastic sheeting. There was no significant difference in the total abundance of spat harvested from the open (X = 0.75 ± 0. 1 1 SE) and bagged (X = 0.61 ± 0.10 SE) collectors (Table 2). When only live spat were considered, there were totals of 24 and 19 spat on open and bagged shademesh. respectively, and 8 spat on both bagged and open collectors made of sheeting. Variation in Collections of P. maculata Among Sites and Collectors More than 1 .000 P. maculata were collected at two sites, and >I00 individuals were collected at another four sites (Table 1). There was a significant positive coirelation (Pearson's r = 0.62, df = 13, p < 0.05) between the abundance of P. margarilifera and P. maculala across all 15 sites. Thus, sites with high abun- dances of P. maculala were relatively good sites for the settlement of P. margarilifera. There was a significant interaction between the effects of site and protection on the abundance of P. maculala (Table 2; p = 0.02). Open collectors had significantly more spat than bagged collectors at two of the three sites (Table 3). There was no signif- TABLE 2. Results of three-v»ay ANOVA for the effects of site (S), material (M), and protection (P) on the total abundance of P. margarilifera and P. maculata." TABLE 3. Mean abundance of P. maculata harvested froin open and bagged collectors at three sites and mean abundance of Cymatium spp. harvested from open and bagged collectors at five sites. Species P maculata Cymatium Site 3 10 12 3 10 12 1 13 Protection' Species Open 43.5 26.0 20.3 Bagged 19 4 P. margarilifera P. maculata Source df F P df F P 8.9 4 1 1 4 4 1 4 180 3.7298 19.0641 0.6652 0.9499 1.4740 2.2250 0.4528 0.0061" 0.0120^ 0.4605 0,4365 0.2119 0,2101 0.7702 1 1 2 2 1 2 4.9544 6,6906 2,4311 1.1269 3.8839 12.5877 0.2550 0.0087" 0.1226 0.2593 0.3278 0.0235" 0.0710 0.7754 37.5 Material 0.80 0.95 Protection S X M S X p 0.45 0.95 2.30 1.45 M X P 0.60 1.55 S X M X P 1.20 1.30 Data for both species were transformed to logm ix + 1). ' Significance: p < 0.01. Significance: p < 0.05. " Underlined means do not differ significantly by SNK test. Real means are displayed, but SNK tests were performed on data transformed to log,o (x + 1). 538 q; o CO •D c CO C CO 0) 35 30 - 25 - 20 - 10 - Friedman and Bell a) 12 10 JO E 6 z 4 I sheeting shademesh sheeting shademesh bagged bagged open open Collectors Figure 2. Mean abundance of P. maculala found on each type of collector. Data were pooled across the three sites of greatest settle- ment. Error bars are standard errors. icant difference in the number off. maculala collected from the two types of material (Table 2; p = 0. 1): a mean of 20.8 ± 5.54 SE spat was collected from shademesh, a mean of 30.9 ± 6.94 SE was harvested from plastic sheeting. Live spat of P. maculata had a higher mean abundance on the open and plastic collectors than on bagged or shademesh collectors (Fig. 2). There was, however, great variation in the number of live spat for each type of collector (Fig. 2). In the case of open col- lectors, for example, 20% of collectors had large numbers of live spat with a mean of 79.6 ± 45.9 SE, whereas the other 80% only averaged 0.9 ± 0.2 SE. Variation in Size of Spat The spat of P. margarilifera alive at harvest ranged from 8 to 71 mm DVM (Fig. 3a). The mean DVM of spat was 32.4 mm (±1.7 SE). All types of collectors had a wide size range of live P. margarilifera. except for the open sheeting collectors, which held none smaller than 30 mm (Fig. 4). The modal size of P. maculata was smaller than that of P margarilifera (Fig. 3b). There were significant differences (df = 3. F = 17.45. p = 0.001) in the size of P. maculala among collectors. P. maculata were significantly larger on open sheeting collectors (X = 22.1 ± 0.59 mm SE) than on_open shademesh (X = 17.9 ± 0.82 mm SE), bagge_d sheeting (X = 16.9 ± 0.55 mm SE), and bagged shademesh (X = 16.3 ± 0.8 mm SE). Ratio of Live Spat to Total Number of Spat There were few differences in the ratios of live to total number of spat of P. margarilifera among the different collectors. The ratios ranged from 40 to 46.3%. The ratios of live to total numbers of spat of P. maculata were higher on open collectors, (68.9% for sheeting and 40. 1% for shademesh) than on the bagged collectors (23.8% for sheeting and 22.9%- for shademesh). P margarilifera n = 70 fl 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 b) JD E z 140 120 100 80 60 40 20 0 P maculata n = 387 5 10 15 20 25 30 35 40 Size class (mm) Figure 3. Size frequency distributions of live spat of P. margarilifera and P. maculala renio\ed from collectors after 6 mo. Data for P. maculata are a subsample of the 1,521 individuals. Variation in Abundance o/Cymatium sp. Among Sites and Collectors We found a total of 185 Cymalium spp. on collectors at the five sites that held the greatest abundance of P. margarilifera. There was a significant interaction between the effects of site and "pro- tection" on the abundance of these gastropods (Table 4; p = 0.047): bagged collectors had significantly more Cymalium spp. than did open collectors at one of the sites (Table 3). At the other four sites, there was no significant difference in the numbers of Cvmaliutn spp. between open and bagged collectors (Table 3). 12 -, u) 10 (0 13 "O > 8 "D C O 6 (/) OJ I ^ 3 0 10 20 30 40+ 0 10 20 30 40+ 0 10 20 30 40+ 0 10 20 30 40+ Size classes (mm) Figure 4. Size frequency distributions (DVM) of live P. margarilifera on the four types of collectors. - Sh 3de op m« en ;sh Shademe bagget ;sh i She op 3tin en g - Sheeting bagged : 1 ~ [— r— 1 1 1 1 Collection of Pearl Oyster Spat 539 TABLE 4. Results of a three-way ANOVA for the effects of site (S), material (M), and protection (Pi on Cymatium spp. abundance among collectors." Cymatium Source Site Material [Protection S X M S X P M X p S X M X 1 Residual df F P 4 0.492 0.745 1 L328 0.313 1 4.674 0.097 4 L007 0.405 4 2.465 0.047" 1 0.963 0.382 4 L628 0.169 180 ■" Data were transformed to logio (x + 1|. '' Significant (p < 0.05). The sizes of Cymalium spp. ranged troni 6 to 67 mm (X = 24.5 ± 0.72 mm SE). There were no significant differences in the mean shell sizes of Cxmatium spp. among the four types of col- lectors . Other Predators Large numbers of predatory portunid crabs. Thalamila quadri- dens (A. Milne Edwards 1869). were found on the collectors. Smaller numbers of xanthid crabs, e.g., Gaillardiellus ohenialis (Odhner. 1925), were also found. These crabs are preda- tor/scavengers and are adapted to scrape off sessile invertebrates (P. Davie, pers. comm). The majority of individuals found had a carapace width <20 mm; however, larger crabs with carapace widths S40 mm were found. The majority of crabs with a cara- pace width >20 mm (76% of all crabs) were associated with collectors ""protected" in mesh bags. Sixteen trigger fish (Balistidae) and three octopus were also collected from the five sites. All but one of these individuals were found on open collectors. DISCUSSION Our study showed conclusively that there were differences in the abundance of both P. inargaritifera and P . macidata with respect to type of collector. These differences appeared to be re- lated to the structure of the collectors, the texture of the materials used in their construction, and the survival rates of spat. Despite the low numbers, P. margiirilifera were more than twice as abundant on shademcsh than on plastic sheeting, being found alive in the pockets and folds of the shademesh collectors at harvest. P. inargaritifera prefer dark surfaces for settlement (Sims 1989), with black or dark blue spat collectors producing the best yields (Coeroli 1983). Both of the substrate we used were black and had the same surface area, but shademesh presented a multi- stranded surface that differed from the smooth surfaces of the plastic sheeting. Dayton et al. ( 1989) postulated that oyster larvae are affected by microscale transport elements and behaviors, such as boundary layers, chemical cues, microtopography. and preda- tion from established individuals. We do not know which charac- teristics of shademesh were favored over plastic sheeting, but the multistrandcd surface evidently provided suitable points for set- tlement. In contrast to P . inargaritifera, abundances of live P. inaculata were far greater on collectors made of plastic sheeting than on shademesh (in the absence of bags). The strips of tlat plastic had a looser structure, allowing spat that had settled to maintain good water exchange. The '"habitat" of the open plastic collectors ap- parently suited P. macidala: individuals found on this type of substratum were significantly larger at harvest than those on the shademesh and closed collectors. Interestingly, the spat of P. mac- idata associated with shademesh were found usually on the edges of the sheets, where they had improved access to water flow. Comparisons of the two materials used to collect spat cannot be made without consideration of the collection environment. Our observations were that, within the high island/open reef environ- ments of Solomon Islands, elevated levels of particulate matter in the water after heavy rainfall, as well as high fouling rates, ac- centuated differences in the "habitats" created by the two sub- strata. These "habitats" influenced the settlement and survival rates of the two pearl oyster species. For example, shademesh was made less suitable for P. inaculata because particulates accumu- lated within "dead spaces." This was not such a problem on collectors made of plastic strips. However, if either type of col- lector became fouled too heavily, or covered in particulates, both species were mostly dead at harvest. Similar problems have been encountered elsewhere. Collections of P. inargaritifera in the Philippines were ""unsuccessful" because of the high productivity of the water and heavy fouling of collectors (Gervis and Sims 1992). Predation plays an important role in the dynamics of pearl oyster stocks (Horncll 1914a & b). The presence of predators on the collectors and the high incidence of dead spat indicate that predation also influenced the number of live spat on our collectors. Our observations indicate, however, that shademesh provided bet- ter protection from predators than the plastic sheeting: the spat of P. inargaritifera were able to live on the inside of folds in the "permeable" shademesh. In this position, spat were presumably afforded greater refuge from predators than on the flat surfaces of the collectors made of plastic sheeting. Judging from the size of the live shells found, P. inargaritifera continued to settle and grow on the shademesh collectors throughout the 6-mo period. In con- trast, small (<30 mm) P . inargaritifera shells were not seen on the open plastic collectors. Ambrose et al. (1992), studying recruit- ment of scallops, proposed that collectors with greater structural complexity may inhibit the activities of predators. For reasons already suggested, P. inaculata were seldom found on the inside folds of shademesh collectors, but they were com- mon on open plastic collectors. In this vulnerable position, how did they persist? The answer may lie in the large number that settled (20% of collectors held over 79 live individuals): predation pressures were msufficient to markedly affect abundances. Although shademesh was two to five times more effective at catching the spat of P. inargaritifera than plastic sheeting, it was four times more expensive. However, this difference in costs is small in proportion to the total farm costs and is outweighed by the fact that shademesh sheets are more easy to handle (thread and shred) than the plastic strip collectors, therefore saving on labor costs. In addition, the shademesh material can be used many times, whereas the plastic sheeting is less easy to recycle. A major finding of our experiment was that the abundance of neither/', inargaritifera nor P. inaculata was significantly greater 540 Friedman and Bell on spat collectors placed inside protective bags. In the case of P. maculata, abundances were actually significantly greater on open collectors than on collectors in bags at two of the three sites an- alyzed in detail. This result can be attributed to two main factors. First, deterioration in the quality of the "habitat" within the bag over the 6-mo period, due to reduced scope for water exchange. This may have been caused m part by the baffling effect of the mesh itself, but was also due to the accumulation of particulates and the heavy fouling of the bags. After 6 mo, the bags supported a complex community of algae, sponges, and ascidians. These organisms blocked much of the mesh. These conditions were likely to lower the settlement success of spat (Suniptonet al. 1990) and reduce their growth (Coeroli et al. 1984). If the bags became fouled heavily or filled with particulates, both species were mostly dead at harvest. Second, the bags were not effective at protecting spat from all types of predation. On the contrary, spat bags included rather than excluded Cymatium spp. and portunid crabs, which are effective predators of juvenile bivalves (Appukuttan 1987. Govan 1994). These animals evidently recruited to the spat collectors as larvae and were trapped within the 2 x 5 mm mesh of the protective bags as they grew. Dayton ct al. (1989) also recognized "extremely strong" predation pressures by invertebrates, e.g.. gastropods, crabs, flat worms, and octopus, on oysters protected from fish predation during a study of oyster settlement on the Great Barrier Reef. In some cases, the "protective" bags were torn, indicating that oysters did not receive comprehensive protection from predatory fish. Chellamet al. (1987) found that they needed to put secondary fish nets (mesh. 10 mm) around spat already protected by fine meshes to stop fish predation during growout. We do not know the relative rates of predation by fish, crabs, and Cymatium spp. and so cannot assess the relative advantage of excluding one type of predator at the expense of retaining another. It is clear, however, that bags do not exclude two important types of predators, Cymatium spp. and crabs, and that, in some cases, the use of bags increases their abundance. CONCLUSIONS The choice of substrate used to construct collectors had a sig- nificant influence on the abundance of spat: P . margaritifera pre- ferred shademesh, and P. maculata preferred plastic sheeting. This implies that further experiments are needed to select the best materials for collecting the spat of P. margaritifera. and that farm- ers may be able to design collectors that target particular species over potentially competitive species. Such experimentation is crit- ical where the collection of spat is marginally effective. Predators such as Cymatium spp. gastropods and portunid crabs settle to spat collectors from the plankton. Bags placed around spat collectors to exclude predators such as fish can enclose Cymatium spp. and crabs as they grow, resulting in increased predation by these invertebrates. "Protective" bags also become heavily fouled. In severe cases, this fouling may render the "habitat" within the bag unsuitable for the growth and survival of pearl oyster spat. Because the number of spat on collectors held in protective mesh bags was significantly lower at some sites and because the installation of bags adds considerably to the cost of spat collectors, we do not recommend the use of spat bags for the collection of pearl oysters within environments similar to those in Solomon Islands. ACKNOWLEDGMENTS We are grateful to M. Gervis and G. Tiroba for their assistance during the study. J. Munro and J. Lucas provided useful comments on the manuscript. This study was conducted with funding from The Australian Centre for International Agricultural Research (ACIAR). LITERATURE CITED Ambrose. W. G. JR.. C. H. Pederston. H. C. Summerson & Junda Lin. 1992. Experimental tests of factors affecting recruitment of bay scal- lops (Argopecten irradians) to spat collectors. Aquacidture 108:67- 86. Appukutlan. K K. 1987. Pearl oyster culture in Vizhinjani Bay. pp. 54— 61. In: K. Alagarswami (ed.). Pearl Culture. Central Marine Fisher- ies Research Institute. Cochin, India. Cabral, O., K Mizuno & A Tuani. 1985. Preliminary data on the spat collection of mother of pearl {Piiictada margaritifera. Bivalve, Mol- lusc) in French Polynesia. Proc. 5lh Int. Coral Reef Congress Tahiti 5:177-182. Chellam. A., T S. Velayudhan & A, C. C, Victor. 1987. Pearl oyster fanning, pp. 72-77. In: K. Alagarswami (ed.). Pearl Culture. Central Marine Fisheries Research Institute. Cochin. India. Coeroli. M. 1983. Pinclada margaritifera. pp. 1-20. In: Milieu lagonaire. Etat des connaissances. Peche Document 1 . E.V.A.A.M., .Service de la Mer, Polynesie Francaise. Coeroh, M., D. De Gaillande, J. P Landret & AQUACORP (D Coala- nea). 1984. Recent innovations in cultivation of molluscs in French Polynesia. Aquaculture 39:45-67. Dayton, P. K.. J. H. Carleton. A. C. Mackley & P. W. Sammarco. 1989. Patterns of settlement, survival and growth of oysters across the Great Barrier Reef. Mar. Ecol. Prog. Ser. 54:75-90, Dharmaraj. S.. A. Chellam & T. S. Velayudhan. 1987. Blofouling. bonng and predation of pearl oysters, pp. 92-97. In: K. Alagarswami (ed.) Pearl Culture Central Marine Fisheries Research Institute. Cochin. India. Gervis, M. & N. A. Sims. 1992. The biology and culture of pearl oysters (Bivalvia: Pteriidae). ICLARM Stud. Rev. 21, ICLARM. Manila, Philippines. 49 pp. Govan, H. 1994. Cymatium muruinum and other ranellid gastropods: ma- jor predators of mancultured tridacmd clams. Thesis submitted for PhD. Heriot-Watt University. Edinburgh. 118 pp. Homell. J. 1914a. An explanation of the irregularly cyclic character of the pearl fisheries. Madras Fisheries Bull. 8:22. Homell. J. 1914b. A preliminary note on the preponderant factor govern- ing the cyclic character of the pearl fisheries of Ceylon and South India, pp. 644—647. In: IX Congress International de Zoologie. Im- primerie Oberthiir. Rennes. France. Intes, A. & M. Coeroli. 1985. Evolution and condition of natural stocks of pearl oysters iPinctada margaritifera Linne) in French Polynesia. Proc. 5th Int. Coral Reef Congress Tahiti 5:545-550. Nicholls. A. G. 1931. On breeding and growth rate of the black-lip pearl oyster {Pinclada margaritifera). Rept. Gt. Barrier Reef Comm. 3:26- 31. Passfield. R. 1989. Basic requirements to set up a small pearl farm . South Pacific Commission Library, Noumea. New Caledonia. 2 pp. Rowntree. J. T. 1993. A preliminary economic assessment of the expan- sion of the Cook Islands cultured black pearl industry: constraints, opportunities and potential impacts. Pacific Islands Marine Resource Collection of Pearl Oyster Spat 541 Project Report 93-01. RDA International Inc. Placerville. USA 149 pp. Shirai. S. 1970. The Story of Pearls. Marine Planning Co.. Okinawa. Japan 132 pp. Sims, N. A. 1988. Pearls ami Pearl Oysters "Poe e Paraii." Cook Is- lands Fisheries Resource Profile No. 2. 10 pp. Sims. N. A. 1989. A literature review: Pinctada mar^arilifera. Submitted in partial fulfillment of a Masters Degree. University of New South Wales. Sims, N. A, 1993. Pearl oysters, pp. 409-403. In: A. Wnght (ed.). Near- shore Resources of the South Pacific. IPS, Suva. 710 pp. Sumpton, W. D., I. W. Brown & M. C. Dredge. 1990. Settlement of bivalve spat on artificial collectors in a subtropical embayment in Queensland, Australia. J. Shellfish Res. 9:227-231. Swift, D. R. 1985. Aquaculture Training Manual. Fishing News Books, Famham, UK. 135 pp. Victor, A. C. C A. Chellam & S. Dharmaraj. 1987. Pearl oyster spat collection, pp. 49-53. In: K. Alagarswami (ed.l. Pearl Culture. Cen- tral Marine Fisheries Research Institute, Cochin, India. J„iii mil ot Shellfish Research. Vol. 15. No, 3. 543-533, 1996. PHYSIOLOGIC VARIABILITY OF EASTERN OYSTERS FROM APALACHICOLA BAY, FLORIDA WILLIAM S. FISHER,' JAMES T. WINSTEAD,' LEAH M. OLIVER,' H. LEE EDMISTON,^ AND GEORGE O. BAILEY^ ^U.S. Environmental Protection Agency Center for Marine & Estiuirine Disease Research Gulf Ecology Division National Health and Environmental Effects Research Laboratory Gulf Breeze. Florida 32561 'Florida Department of Environmental Protection Apalachicola National Estuarine Research Reserve Apalachicola, Florida 32320 ABSTRACT Eastern oysters. Crussosliea virgiiiicu. were collected monthly during a I -y period from two study sites in Apalachicola Bay. FL. and several measurements were made of their physiologic condition. Continuous and intermittent temperature measurements at both sites shows highly coincident ambient temperature regimens. Salinity measurements, however, were erratic and varied dramatically between sites. Oyster gonad size and gametogenic condition were highly synchronous at both sites, supporting the concept of temperature-driven reproductive cycles. Other measurements, including condition index. wet:dry tissue weight ratio, digestive tubule condition, and vesicular connective tissue condition, showed significant variability as the result of sampling month, but also differed because of site and/or interaction between date and site, indicating that local effects influenced oyster physiology. Temperature control over condition index and wet:dry tissue weight seems apparent, but it is not known whether the changes resulted directly from temperature or from temperature-driven reproductive and metabolic cycles. A significant difference between site means at specific dates was observed for digestive tubule condition and may relate to short-term salinity differences. Other physiologic variations could not be attributed to any of the physical conditions monitored (temperature, salinity, pH, and dissolved oxygen). The variability of oyster physiologic measurements inherent at different sites and seasons must be well understood to properly interpret them in the context of biologic indicators of environmental condition. KEY WORDS: Eastern oysters. Crussoslreu vir^inicu. bivalve physiology INTRODUCTION The interpretation of physiologic measurements of bivalve molluscs as indicators of their health or of environmental condition is complicated by the wide ranges and multiple sources of vari- ability. A primary source of variability is the reproductive cycle, which is driven by temperature and causes annual metabolic changes associated with gonadal development and spawning (Galt- soff 1964, Quick 1971). Other environmental factors in oyster habitats, such as salinity, dissolved oxygen, nutrients, toxicants, parasites, and disease, typically show intermittent and short-term fluctuations, and their effects can be either obscured or magnified by the seasonal reproductive effects. The proper interpretation of physiologic measurements, whether used as indicators of health (Engle 1951; Bayne et al. 1976) or as biologic indicators of en- vironmental condition (e.g.. International Mussel Watch [IMW| 1980), requires a concrete understanding of both seasonal and intermittent variations in the natural environment. This study describes the range of variability in certain physio- logic measurements of eastern oysters, Crassostrea virginica, in a Gulf of Mexico estuary and attempts to distinguish seasonal re- productive effects from those caused by local environmental fluc- tuations. Two oyster beds were selected in Apalachicola Bay (Franklin County, FL) to study monthly changes in selected phys- iologic and immunologic characteristics. The sites lie within 15 km of each other in a bay system protected by a chain of barrier islands (Gorsline 1963). Their proximity assures a reasonably sim- ilar temperature regimen, whereas currents, tides, and river inflow create smaller-scale and intermittent environmental diversity, which can cause physiologic differences between inhabitants of the two sites. The Apalachicola Bay system (Fig. 1 ) is a wide (210 square miles), shallow, and relatively unpolluted body of water (Thomp- son et al. 1990) fed primarily by the Apalachicola River. The system is a highly productive lagoon/barrier island complex that typically yields $12-$ 16 million in dockside seafood landings an- nually (Florida Department of Natural Resources 1986). Since the early 190()s, commercial fishing has been the most important eco- nomic activity within the bay and currently supports 60-85% of the local community (Rockwood and Leitman 1977). Historically, revenue from this industry has accounted for nearly half of Frank- lin County's income (Whitfield and Beaumariage 1977). The eastern oyster is the most important commercial inverte- brate in Apalachicola Bay, which supplies nearly 90% of the oys- ters harvested in Florida. Production on commercial oyster bars has been estimated at between 400 and 1 .200 bushels/acre per year (Ednoff 1984. Berrigan pers. comm.). Because of relatively mild temperatures in the area, rapid oyster growth is sustained through- out the year. Harvestable oysters, those larger than 3 inches (76 mm), have been grown from spat in as little as 39 wk. The spawn- ing season is one of the longest in the United States (Ingle and Dawson 1952). Although oysters from Apalachicola Bay have frequently been evaluated from a commercial standpoint, they have not been ex- tensively investigated in terms of their physiology, immunology, and disease. The monthly measurement of several physiologic 543 544 Fisher et al. Figure 1. The Apalachicola River and Bay system with the two oyster collection sites. Cat Point Bar and Dry Bar (also known as St. Vincent's Bar), noted with stars. The sites are in northwest Florida (inset). characteristics presented here provides a baseline of infonnation on comparatively vigorous and stable oyster populations in a rel- atively unpolluted environment. MATERIALS AND METHODS Collection Sites Oysters were collected from two of the most commercially productive oyster bars in Apalachicola Bay — Cat Point and Dry Bar (also named Drybar and St. Vincent's Bar). The Apalachicola estuarine system is divisible into four sections based on natural bathymetry and man-made structural alterations: East Bay. St. Vincent Sound, Apalachicola Bay. and St. George Sound (Fig. 1). The average depth in these bays ranges from 1 m in East Bay to 3 m in Apalachicola Bay. with maximum depths of 7 m occurring near the barrier islands (Dawson 1955, Gorsline 1963). The depths of the two oyster bars ranged from I to 2m. Oyster bars cover over 10,600 acres, or —10% of the sub- merged bottom within the boundaries of the Apalachicola National Estuarine Research Reserve (NERR). The two oyster collection sites are within NERR and were selected for their relatively large, easily harvested oysters, their locations on either side of the Apalachicola River (Fig. I ). and the availability of historical water quality data from the Florida Department of Environmental Pro- tection (Thompson et al. 1990). The Dry Bar site is off St. Vin- cent's Island on the western edge of Apalachicola Bay. ~8 km WSW from the mouth of the Apalachicola River (latitude [Eat] 29:40.25, longitude [Eon] 85:03.33). The Cat Point site is located on the western edge of St. George Sound, just east of the St. George Island bridge, ~6 km ESE of the mouth of the Apalach- icola River (Lat 29:43.05, Eon 84:52.93). The two sites are less than 15 km apart. Both sites lie within the chain of protective barrier islands, but the physical and chemical characteristics of the water are influenced by wind, tide, and river currents, as noted above. Water Quality Measurements Water salinity and temperature were recorded from both sites on each oyster collection date with a refractometer and digital thermometer. All ""continuous" water quality characteristics were recorded every 30 min from June 1992 to December 1993. except for short maintenance periods. They were measured with Hydro- lab' Datasonde 3 dataloggers that were programmed to measure temperature (±0.15 C), pH (±0.2 units), dissolved oxygen (±0.2 mg/L). salinity (±0.2 ppt), and tide height (±0.09 m) (Hydrolab Corporation 1991 ). The dataloggers were equipped with an exter- nal submersible battery pack and 70K extended memory, which allowed in situ monitoring for extended periods. Divers attached the datalogger units to pilings located at the sampling sites with two stainless steel clamps and a safety cable. The probes were situated 0.4 m above the bottom substrate at both collection sites. The units were retrieved every 14-21 days (de- 'The mention of commercial products does not constitute endorsement by the U.S. Environmental Protection Agency. Physiologic Variability of Eastern Oysters 545 pending on fouling) for downloading, cleaning, and rccalibration. At this time, all electrolytes were replaced as well as the dissolved oxygen low-flow membrane. Laboratory-grade standards were used to calibrate the instruments for salinity and pH. Air calibra- tion was used for the dissolved oxygen probe. Salinity readings remained within ±0.5 ppt of the calibration standard at the end of each sampling period, and pH readings also showed very little deviation from calibration, with a maximum error of only ±0.1 units. The dissolved oxygen measurements showed the greatest drift in calibration, up to several milligrams per liter. This was primarily due to fouling of the low-flow mem- brane by algae, silt, and barnacle settlement. In an effort to rem- edy this situation, a plastic mesh screen was placed around the probe guard to reduce fouling while still allowing water to flow past the probes; also, the sampling period was reduced to 14 days during the warmer months when fouling was greater. Oyster Collection and Processing Oysters were collected on (approximately) a monthly basis with hand tongs from October 1991 to October 1992, and addi- tional samples were collected in December 1992 and March 1993. Dates of collection were (1991): October 22, November 19. De- cember 19; (1992): January 21, February 25, March 31, April 21, May 19, June 23, July 21, September I, September 29, October 26, December 7; and ( 1993): March 29 for a total of 15 collections over a span of 17 mo (525 days). The September 1 sample is referred to as the August sample in this report. Oysters were se- lected at an approximate 80-mm height (umbo to edge of bill) and 50-mm length (greatest distance across the shell orthagonal to height) and were placed immediately into coolers containing cold ice packs. The coolers were transported to the Environmental Pro- tection Agency Gulf Ecology Division at Gulf Breeze, FL, and were placed in a refrigerator (4°C) overnight. Twelve oysters from each site were cleaned of fouling organ- isms and held at room temperature for 1-2 h. After the total oyster volume was measured, as described below, the shells were notched with a grinder at the margin of the shell and hemolymph was withdrawn from the adductor muscle with a syringe and 22- gauge needle. Hemolymph was used to measure Perkinsus mari- nus infection intensity, described here, and several defense-related characteristics that are reported elsewhere (Fisher et al. 1996). Condition Index and Wel.Dry Weight Ratio The total volume of each oyster was estimated by weighing (±2 g) water displaced from a beaker when the oyster was added (assuming I g = 1 niL of water). Weighing the displaced water provided greater sensitivity than directly measuring the volume of displaced water. Estimates of shell volume after the removal of all soft tissues were obtained in the same manner, and the difference was calculated as an estimate of the internal shell cavity volume. Soft tissues removed from the shells were blotted and weighed to obtain the total wet weight and then were dissected for histology (see below). The remaining tissues were weighed (partial wet weight), dried at 60°C for 48 h, and then reweighed (partial dry weight). The resulting wet:dry ratio was used to estimate the total dry tissue weight. The condition index (CI) was calculated as (Galtsoff 1964): Gender and Gonadal Condition For histologic sectioning, a 3- to 5-mm-thick band of tissue was cut transversely with a razor blade at a distance of one-fourth to one-third of the length of the animal from the umbo in such a manner as to contain portions of mantle, gill, digestive tubule, and gonad. By standard histologic procedures, the dissected tissue was fixed for 24-72 h in Davidson's (formalin) fixative and stored in 70% ethanol before paraffin embedding. After embedding and sectioning with a microtome, slides were stained with hematoxylin and eosin. Gonad portions of the slide were examined with light microscopy to determine gender and were graded for gonadal con- dition (GC) (Table 1), which was adapted from measurements made by the International Mussel Watch Program (IMW 1980). Relative Gonad Size The greatest diameter of adductor muscle was measured with Fowler digital calipers while one end of the muscle was still at- tached to the half-shell. Gonadal thickness was measured at a consistent site opposite the gills on the band of tissue isolated for histologic sectioning. Gonadal material external to the digestive tubules was measured. These measurements were used to calculate the relative gonad size (Galtsoff 1964): RGS gonadal thickness (mm) x 100 diameter of adductor muscle (mm) Structure of Digestive Gland and Connective Tissue Histologic slides were examined (400 x magnification) for the structure of digestive diverticulae and vesicular connective tissues. Tubules of digestive diverticulae were measured by the technique of Winstead (1995), with a light microscope (200x) equipped with an ocular micrometer. The area of the section containing digestive tubules was divided into four quadrants, and five tubules from each quadrant were measured. Two sets of measurements TABLE 1. A description of gametogenic characteristics observed in histologic sections of oyster gonads and the values assigned for the determination of GC. Value Observations CI = total dry tissue wt (g) x 100 internal shell cavity volume (mL) 0 Neuter or resting stage with no visible signs of gametes 1 Gametogenesis has begun with no mature gametes 2 First appearance of mature gametes to approximataely one-third mature gametes in follicles 3 Follicles have approximately equal proportions of mature and developing gametes 4 Gametogenesis progressing, but follicles dominated by maUire gametes 5 Follicles distended and filled with npe gametes, limited gametogenesis. ova compacted into polygonal configurations, and sperm have visible tails 6 Active emission (spawning) occurring; general reduction in sperm density or morphological rounding of ova 7 Follicles one-half depleted of mature gametes 8 Gonadal area is reduced, follicles two-thirds depleted of mature gametes 9 Only residual gametes remam. some cytolysis evident 10 Gonads completely devoid of gametes, and cytolysis is ongoing 546 Fisher et al. were taken from each tubule — a lumen diameter and total tubule diameter — and from these values, a tubule ratio was calculated (lumen diameter/tubule diameter), with increasing values indicat- ing more squamous digestive tubule cells. An average digestive tubule ratio (DTR) for an oyster was obtained from the 20 mea- sured tubules. The tubules also were ranked subjectively according to epithe- lial layer condition, as judged by cell size and morphology: I = large and columnar. 2 = average, and 3 = small and squamous; the morphology of the tubule cells determines the thickness of the epithelial layer and the size of the lumen. These vary within an individual oyster but were sufficiently consistent to be easily clas- sified in this scheme. An earlier study indicated that these two techniques, the quantitative and the semiquantitative, produced similar results (Winstead 1995). Vesicular connective tissue (VCT) was examined for structural and morphological properties that were graded as 1 = lacy, intact, and regular network (orga- nized), 2 = moderate edema and tearing with slight hemocyte response, or 3 = extensive edema and tearing with heavy hemo- cyte response. Parasites and Disease Examinations were made of histologic sections to assess the type and relative intensity of parasitic infections and noninfectious diseases or abnormalities. Each oyster was examined for Bucepha- alus spp. and Proctoeces spp. (digenetic trematodes), Nemalopsis spp. and P. marinus (protists). Tylocephahim spp. (cestode). rick- ettsial inclusion bodies, and thigmotrichous ciliatc protozoans, as well as noninfectious neoplasias and hemocytic or inflammatory responses. Hemolymph Diagnosis of P. marinus The prevalence and infection intensity of the protozoan parasite P. marinus was determined by a modification of the method of Gauthier and Fisher (1990). Hemolymph drawn (0.5 niL) from the adductor muscles of individual oysters was centrifuged for 4-5 min at 2.940 x g in a microcentrifuge. The supemate (cell-free hemolymph. or serum) was collected and held at 4°C for analyses described in a separate article (Fisher et al. 1996). The pelleted hemocytes were covered with 0.5 mL of Ray's fluid thioglycollate medium to which 2.5 jxL of Chloromycetin stock solution was added and mixed; then, 100 \xL of mycostatin stock solution was layered on top (Ray 1966). These were incubated in the dark for 5-7 days then were centrifuged as described above, incubated for 1 h at 60°C in the presence of 2 M sodium hydroxide, washed twice by centrifugation and suspension in distilled water, and fi- nally resuspended in 0.5 mL of Lugol's iodine solution. Samples were filtered onto 0.22-(j.m filter paper for microscopic examina- tion ( lOQx ) and enumeration. Statistical Methods Data were entered into SAS (SAS Inc.. Gary. NG) and ana- lyzed by use of the General Linear Models and the Shapiro-Wilk test for normality. Residual plots were examined to assure homo- geneity of variance and independence and normality of error terms in the resulting models. P. marinus and wet:dry tissue weight data were log,,, transformed to achieve compliance with assumptions of analysis of variance (ANOVA). Two-way analysis of variance (ANOVA) was conducted to relate each dependent variable (phys- iologic measurements) to the main effects of date and site and to test for possible interactions between date and site. The results of all analyses are reported, but main effects are discussed only if there was no significant interaction. Where significant main ef- fects were found. Tukey's post-hoc test was used to differentiate between overall date or site means. For variables that displayed a significant interaction effect, differences due to date for each of the sites were examined with Tukey"s post-hoc results, which compared means of all date * site combinations. Gorrelational analysis (Pearson's procedure) was used to relate collection date salinities with digestive tubule condition and wet:dry tissue weight measurements and to compare qualitative and quantitative (DTR) measures of digestive tubule condition. Levels of significance and high significance were selected by convention as p =s 0.05 and p =s 0.01. RESULTS Temperature Variations Water temperatures were very consistent between the two col- lection sites, both for temperatures recorded at the time of collec- tion (Fig. 2. top; October 1991 to March 1993) and during the continuous monitoring phase (June 1992 to March 1993). as dem- onstrated by the daily averages (Fig. 2, bottom). Daily averages rarely varied more than 3°G between sites. Both sites exhibited a seasonal cycle with progressive warming after January and cooling after July. Examination of 30-min data 7 days before monthly collection of oysters (June to December only, not shown) did not reveal any conspicuous fluctuations. Salinity Variations Salinity measured at the time of collection was the same for both sites at only 3 of the 15 collection dates (Fig. 3. top). These measurements displayed an inconsistent profile, with highest sa- linities reaching 25 ppt and lowest salinities reaching as low as 5 ppt for both Gat Point Bar (February 1992) and Dry Bar (March and December 1992). Neither site exhibited a strong seasonal pat- tern, although there appeared to be a low-salinity period during winter to spring (January to March). Gontinuous salinity monitoring from June 1992 to March 1993 revealed striking differences between sites in daily averages (Fig. 3, bottom); differences were often 10-15 ppt and were sometimes as high as 25 ppt (January 1993). During the 10-mo period, sa- linities ranged between 35 and 1 ppt. Examination of the 30-min sampling data (not averaged) for both sites revealed large varia- tions in salinity, sometimes varying 10-15 ppt within 1 h. Both sites also exhibited periods of stable salinities, often for 2 wk or longer. These data make clear that salinity measured only at the time of collection can be misleading. For example. —20 ppt sa- linity was recorded at the time of collection for Dry Bar in June, even though salinities predominantly ranged from 25 to 35 ppt during this month. Examination of 30-min data 7 days before each monthly col- lection of oysters revealed a notable difference between the two sites in December 1992. Before the collection date (December 7), Gat Point fluctuated between 10 and 25 ppt (Fig. 4). During the same 7-day period. Dry Bar salinity fluctuated between 10 and 30 ppt and then declined to 5 ppt by December 7 (Fig. 4). Dissolved Oxygen, pH, and Tidal Height During the June 1992 to March 1993 continuous monitoring period, the daily average dissolved oxygen concentration gradu- Physiologic Variability of Eastern Oysters 547 o o E M A M J J Date Cat Point Bar ^ — Dry Bar Dry Bar Cal Potnt Dry Bar Figure 2. Water temperature of Cat Point Bar and Dry Bar during the period of collection. (Top) Temperatures taken at the time of collection throughout the entire study period. (Bottom! Daily averages of temperatures taken every 30 min from June 1991 to March 1993 only (note scale differences). All temperatures were recorded less than 0.5 m from the bav bottom. Cat Point Dry Bar Figure 3. Water salinity of Cat Point Bar and Dry Bar during the period of collection. (Top) Salinities measured at the time of collection throughout the study period. (Bottom) Daily averages of salinities measured every 30 min from June to March 1993 only (note scale differences). All salinities were recorded less than 0.5 m from the bottom of the bav. ally rose from 4 to 6 mg/L. Similar patterns were exhibited at both sites. The measurement of pH fluctuated from 7.6 to 8.2 during this period, and there were some differences between the two sites. From June to September 1992. the pH at Cat Point remained near 7.8, whereas the Dry Bar pH fluctuated around 8.0. Both sites decreased to 7.7 or below in January 1993 but returned to above 8.0 by February. The tidal height recorded during this period was very similar at both sites. Tide levels fluctuated near 1 .4 m in June and gradually declined to 1.2 m by March. Relative Gonad Size Average gonad size relative to adductor muscle diameter (Fig. 5)" exhibited a significant date * site mteraction (Table 2). The significant interaction appeared to be primarily a function of rel- atively low variability about the mean, because both sites followed "Data are presented in graphic form tor convenience; tabular data sets are available upon request to the authors. a similar temporal pattern (Fig. 5). When analyzed by post-hoc analysis, relative gonad size (RGS) was largest in April 1992 and smallest in October 1991 and September to December 1992 at both sites (Table 3B). Oysters of similar size were collected on all dates, so adductor muscle average diameter ranged only from 12.8 to 18.0 mm for oysters at Cat Point and 16.3 to 19.6 mm for oysters at Dry Bar. Gonad thickness, however, increased from 1.9 (October 1991 ) to 1 1 .6 mm (April 1992) for Cat Point oysters and from 1.9 (October 1991) to 14.2 mm (April 1992) for Dry Bar oysters. Thus, most of the change in RGS was due to gonadal development rather than changes in adductor muscle size. A sig- nificant difference between sites was detected only in November 1991 (Table 38). Gender and GC Histologically rated GC (Fig. 6) exhibited a significant date * site interaction (Table 2), although GC also demonstrated a high degree of synchrony. Site means were significantly different only during May, August, and September 1992 and March 1993 548 Fisher et al. 35 -r 11/30 12/01 12/02 12/03 12/04 DATE 12/05 12/06 12/07 Cat Point Dry Bar Figure 4. Water salinity of Cat Point Bar (solid line) and Dry Bar (dotted line) measured every 30 min during the week before the December 7 (1992) collection of oysters. Dry Bar salinity dropped to ~5 ppt 2 days before collection, whereas Cat Point Bar salinity remained near ~15 ppt. (Table 3B). The significant interaction appeared to be primarily a function of low variability about the mean, because both sites followed a similar temporal pattern. Gametogenesis (histologic rating = 1) was initiated by February; spawning ( =5) began by May and was nearly complete ( = 10) by October at both sites (Fig. 6). There was no evidence of gametes during November 1991 to January 1992. However, in the following year, gametogenesis was observed in oysters from both sites by December. Oysters were essentially neuter (no gametes visible) from Oc- tober 1991 to January 1992 (Fig. 7). At the onset of gametogenesis (February and March 1992), approximately half of the oysters were males and half were females. The ratio became predomi- o □ 9 8 Cat Point Bar Dry Bar — ^- y/- ^Mf^ffii^m^^^" # Figure S. Average RGS measured during October 1991 through March 1993 from oysters collected at Cat Point Bar and Dry Bar (n = 12). Calculation of RGS was the ratio of gonad thickness to adductor muscle diameter x 100. ND, no data; bars indicate standard errors. TABLE 2. Results of an interactive ANOVA with date and site as independent variables." Variable Date (p<) Site(p<) Datc*Site (p<) RGS 0.001" 0.072 NS' O.OO,^'' GC 0.001" 0.178 NS' 0.001" CI 0.001" 0.004" 0.089 NS' Wet:dry weight (log,o) 0.001" 0.001" 0.001" DTR 0.001" 0.299 NS^^ 0.001" VCT 0.001" 0.001" 0.007" Intensity of P . marlniis infection 0.001" 0.551 NS' 0.069 NS' " All tests were performed from October 1991 through March 1993, except for the diagnosis of P. marinus. which was started in December 1991. " Highly significant difference (p =s 0.01). ' NS, no significant difference. '' Significant difference (p «£ 0.05). Physiologic Variability of Eastern Oysters 549 TABLE i. Study results. A B RGS GC W:D DTR VCT Date CI P. mar CP DB CP DB CP DB CP DB CP DB Oct 91 E — IJ J A A CDEFG CDEFG ABC ABCD ABCD BCD Nov ABCD — EFGH BC K K EFGHI GHI DEF EF AB CD Dec ABCDE ABCD BCD BCDE K K EFGHI GHI EF EF CD CD Jan "92 ABCD A BC BCDE K K GHI EFHGI DEF EF CD CD Feb ABCDE ABCD B BC JK JK FGHI CDEFG F A CD D Mar ABC D B B HI 1 DEFGH HI F F CD D Apr AB A A A GH G GHI HI CDEF EF BCD CD May ABCDE ABCD BCDE B CD EF EFGHI EFGHI F EF BCD BCD Jun ABCD CD CDEF CDEF DEE CDE EFGHI GHI EF DEF BCD ABCD Jul A AB BCDEF BCDEF CDEF BC EFGHI EFGHI BCDEF BCDE ABCD BCD Aug DE CD FGHI CDEF C F AB BCDEF AB BCDEF CD CD Sep CDE ABCD HIJ HIJ C AB BCDEF A CDEF EF A AB Oct BCDE AB HIJ IJ A A ABCD EFGHI ABCD EF ABC CD Dec CDE ABC IJ GHIJ JK K ABCD ABC EF ABCD CD CD Mar '93 ABCDE BCD DEFG CDEF J 1 BCDE GHI DEF EF CD CD ' (A) Significant differences due to sampling dale (Tukey's procedure). Matching letters in a single column indicate no significant differences between collection dates for that variable. Analysis was performed with site data combined for variables with no significant date * site interaction (see Table 2). (B) Significant differences between all date * site means. Matching letters in a single column indicate no significant differences. W:D. logn, wet:dry weight; P mar., level of infection by P marinus in hcmolymph. nantly female during spawning (April to September), with some collections yielding 92% female oysters (Fig. 7). Oyster Condition Index and Wet:Dry Weight Ratio Average condition index varied significantly over collection date and site (Table 2) and had no significant date + site interac- tion. Values were higher for Dry Bar oysters (Fig. 8). which peaked in July and declined rapidly during the next 2 mo (Table 3A). Oyster tissue wet:dry weight ratios (Fig. 9) exhibited a sig- nificant date * site interaction (Table 2). Post-hoc analysis found that wet:dry weight ratios were highest during August to Decem- ber 1992 for Cat Point Bar and in September for Dry Bar (Tabic 38) and that significant differences between sites occurred in Sep- tember and October 1992 and March 1993. Structure of Digestive Gland and Connective Tissue Results from the subjective and quantitative measures of di- gestive tubule condition were strongly correlated; Pearson corre- lations between the two tests for all samples were highly signifi- cant (r = 0.864. n = 357). The quantitative DTR (Fig. 10) was found to have significant date * site interaction (Table 2). Post- hoc analysis found significant site differences in February. Octo- ber, and December 1992 (Table 38). The February 1992 value was the highest for all oysters in the study and contrasted sharply with the low values in preceding and succeeding months at both sites (Fig. 10). Tubule ratios measured from Dry Bar oysters in October and December 1992 were different than those recorded in October and December 1991. respectively. Subjective ratings of VCT structure (Fig. 1 1) exhibited signif- Q z o o _j < < O o Figure 6. Average GC measured during October 1991 through March 1993 from oysters collected at Cat Point Bar and Dry Bar (n = 12). The assignment of values to gametogenic stages is described in Table 1. ND, no data; bars indicate standard errors. n NEUTER ■ MALE □ FEMALE '^ S^ ^^ ^ ^<^ / Figure 7. Gender frequency of 12 oysters collected at Cat Point Bar (left) and Dry Bar (right) during October 1991 through March 1993. Gender was determined by microscopic examination of histologic sec- tions. ND, no data. 550 Fisher et al. X LU Q o O o^>V%^>V//.^^^>%<^VV^^^ / Figure 8. Average CI measured during October 1991 thruugh March 1993 from oysters collected at Cat Point Bar and Dry Bar (n = 12). The calculation of CI was the ratio of dry tissue weight to estimated internal shell cavity volume x 100. ND, no data; bars indicate stan- dard errors. icant date * site interaction (Table 2). Post-hoc analysis found significant differences between sites in November 1991 only (Ta- ble 3B). The lowest VCT values (best condition) occurred during December to March, and the highest levels occurred during Sep- tember at both sites (Fig. 1 1). Parasites and Disease Both sites showed high prevalences of Nematopsis spp.. Tylo- cephalum spp., and P moninis. Nematopsis spp. occurred in >75% of the oysters throughout most of the study period, with exceptions in April (66%) and September (58%) at Dry Bar. In- fection by Tylocephatum spp. was much more variable. At Cat Point Bar, the highest prevalences (100%) occurred in July and September and the lowest (25%) occurred in March and May. At Dry Bar, the highest prevalences (83%) were found in December, May. June, and October and the lowest (17%) occurred in April. There were no apparent temporal trends or site differences in the prevalences of these parasites. Hemolymph diagnosis of P. marinus found 100% prevalence at both sites throughout the study period, with infection intensities generally less than 1,000/niL of hemolymph (Fig. 12). Infection cc < Q 3 m = t UJ s < a z UJ 0.7 n ' Cat Point Bar — ■— Dry Bar -^4^ -^^ 0.6 - f 0.5 - 1 ! f\ i ] ]- 0 4 - l l\ A^ i/ \ 0.3 - J \ i\i J ^i) f Y \ , 0.2 - mvA J ^ K /\ A 0.1 - Y^^ \^ t-' -" ^ a ' i n n ^ z // 1 1 1 1 1 1 1 1 1 1 1 // 1 oC^,^0<3<^Vr^^x Figure 10. Average ratio of digestive tubule lumen diameter to outer diameter measured during October 1991 through March 1993 from oysters collected at Cat Point Bar and Dry Bar (n = 12). Measure- ments were made with an ocular micrometer on histologic sections of oyster digestive diverticulae. ND, no data; bars indicate standard er- rors. intensities fluctuated erratically, and significant differences due to collection date were found; however, no significant effects attrib- utable to site or date * site interaction were seen (Table 2). Sig- nificant increases and decreases were found between collection dates throughout the sampling (Table 3A). DISCUSSION Water temperatures during the study period exhibited a sea- sonal cycle that was highly coincident at Dry Bar and Cat Point Bar (Fig. 2). Perhaps as a consequence, the reproductive status (RGS and GC) of oysters was relatively synchronous between the two collection sites. This was demonstrated by the lack of signif- icant difference due to site (Table 2), although there was signifi- cant date * site interaction. There is evidence that eastern oysters in different locations can have genetically determined environmen- tal requirements for gonadal maturation (Loosanoff 1969, Barber et al. 1991) and contrasting evidence that they are controlled pri- marily by temperature (Ruddy et al. 1975, Butler 1955). In this study, any genetic differences that might exist between oysters at UJ > cr Q t-' UJ 12.5 - Cat Point Bar — ■-- Dry Bar — ^ ~^'\ 10.0 - Wi^t/^ >^ i , \ 7.5 - V 5 0 - 2.5 - Q Z II 1 1 1 1 1 1 1 // 1 ooV>V>>%^V#.^>%<^V>V ^^ Figure 9. Average ratio of tissue wet weight to dry weight measured durini> October 1991 through March 1993 from oy.sters collected at Cat Point Bar and Dry Bar (n = 12). ND, no data; bars indicate standard errors. 3 n Cat Point Bar — ■— Dry Bar - ■^ '\ O 2 - < cr t- o 6s^ b. h^ > 1 - » •=#=» 0 - 1 1 1 1 1 1 1 1 1 1 1 I Q Z 1 rV.^^ Figure 11. Average rating for VCT structure examined in histologic sections during October 1991 through March 1993 from oysters col- lected at Cat Point Bar and Dry Bar (n = 12). ND, no data; bars indicate standard errors. Physiologic Variability of Eastern Oysters 551 X Q- > o UJ <0 3 .C ra E a; w o o 4 ^ Cat Point Bar -m— Dry Bar — ^ ^7^ 1 3 - vK Ai 1 T/t^^ k 2 - /l^ 1^ *r^ 1 - T 0 - 1 1 1 1 1 1 1 1 Q 1 1 1 1 r^^^ o^VA^'^//#'.^>Vo^VV Figure 12. Average of logm-transfornied estimates of F. inarinus in- tensity in 1 ml. of oyster hemolymph. Oysters were collected at Cat Point Bar and Dry Bar (n = 12) during October 1991 through March 1993. Protozoan meronts were counted after culture in fluid thio- glycollate medium and staining with Lugol's iodine solution. ND, no data: bars indicate standard errors. Cat Point and Dry Bar did not appear to interfere with their re- productive synchrony (Figs. 5 and 6). In contrast to temperature, salinities at the two sites were quite different. Data recorded at the time of collection (Fig. 3, top) found salinities to vary as much as 10 ppt between sites even though there was a general pattern of low salinity at both sites during winter to spring and high salinity during fall. Salinity data from continuous monitors deployed from June 1992 to March 1993 (Fig. 3. bottom) showed dramatic fluctuations and differ- ences between sites. These differences occur because of various influences from river flow, tidal action, and wind. The Apalach- icola Bay system is an area of mixed tides, ranging from 0.3 to 0.7 m. that alternate between the semidiurnal tides of southwestern Florida and the diurnal tides of northwestern Florida (Dawson 1955, Gorsline 1963). Net water movement in the system is gen- erally east to west (Ingle and Dawson 1953, Conner et al. 1982), with currents governed primarily by the astronomical tides. These can be influenced, however, by strong prevailing winds (Esta- brook 1973). The Apalachicola River is the primary source of fresh water in the bay. carrying an annual mean discharge of approximately 15,000 (range, 9,300-80,000) cubic feet per sec- ond (U.S. Army Corps of Engineers 1978). Because of these factors, river flow dominates salinities at Dry Bar whereas tidal exchange may equal the river's influence at Cat Point Bar. During periods of light wind or an east wind, the influence of the river was primarily toward the southwest and Dry Bar. Winds from the west shifted the river's influence so that fresher water was found at Cat Point Bar. Local rainfall during the study period did not appear to have an immediate effect on salinity at either site. The wet:dry tissue weight ratio (Fig. 9) of oysters increased near the end of spawning (Fig. 8) and was generally the inverse of the CI. Both characteristics were probably associated with de- graded oyster condition typical for oysters during this period (So- niat and Ray 1985). The lower CI in autumn for Apalachicola Bay oysters is similar to that reported for eastern oysters from some Virginia estuaries (Chesapeake Bay; Austin et al. 1993) and from Galveston Bay, TX (Soniat and Ray 1985), but contrasts with other studies in Chesapeake Bay (Engle 1951) and Louisiana (Hopkins et al. 1954) that have shown CI to be low in summer and high during fall and winter. This apparent lack of synchrony among different oyster populations, even when the sites are rela- tively close, may be related to the availability or quality of nutri- ents (Soniat and Ray 1985, Austin et al. 1993). If so. the utility of CI as a biologic indicator (Scott and Middaugh 1978, Scott and Vemberg 1979, Lawrence and Scott 1982, Roper etal. 1991) must be limited to situations where nutrient input was closely moni- tored. Nutrient differences may also explain why Dry Bar oysters had a significantly higher CI than Cat Point Bar oysters throughout the study period. Digestive tubule atrophy in bivalves has been associated with xcnobiotic exposure (Lowe et al. 1972, Bayne et al. 1976. Chagot et al. 1990, Weis et al. 1995) and environmental stressors such as temperature, spawning, salinity, or nutrition (Thompson et al. 1974, Bayne et al. 1981. Couch 1985, Widdows and Johnson 1988). Stress, regardless of the source, appears to cause epithelial atrophy through the formation of autolysosomes in tubule cells. In this study, the condition of digestive tubules (DTR; Fig. 10) may have been influenced by temperature or temperature-driven phys- iologic changes because tubule atrophy (high DTR) was greater toward the end of active spawning, at least for Cat Point Bar oysters (Table 3B). Conspicuously high DTR were found in Feb- ruary and December 1992 for Dry Bar oysters (Fig. 10) and may have been related to acute salinity changes. Salinities at Dry Bar were low at those collection dates (Fig. 3, top), and data from the continuous water monitors showed a rapid decrease just before the December 7 sampling date (Fig. 4). An acute decrease in salinity may have caused oysters, which are osmoconformers, to close their shells and stop feeding. Winstead (1995) has recently dem- onstrated that digestive tubules in unfed oysters can significantly atrophy, compared with fed controls, within 2 days after the ces- sation of feeding. Alternatively, changes in phytoplankton quan- tity or quality could create the same effect and be related to salinity changes (e.g., river flow). Regardless, factors in digestive tubule variability cannot be accurately assessed by monthly monitoring if changes occur as rapidly as indicated by Winstead (1995). VCT structure appeared to relate to the changing condition of the oyster reproductive cycle, showing the poorest structure (high- est ratings) near the end of spawning, with recovery by early winter (Fig. 1 1). A significant difference between sites was found in November 1991, when Cat Point Bar oysters had poor connec- tive tissue structure relative to Dry Bar oysters. This could be related to the smaller gonad size for Cat Point Bar oysters on that date (Fig. 5), but the factors responsible are unknown. The low P . marinus infection intensities observed in Cat Point Bar oyster hemolymph in March 1992 (Fig. 12) coincided with low salinity for that date and site (Fig. 3, top). This is relevant because P. marinus require salinities above 6 ppt to sporulate (Perkins 1966. Chu and Greene 1989) and transmission in nature is reduced at salinities below 12 ppt (Andrews and Hewatt 1957, Paynter and Burreson 1991). However, the salinity was even lower at Dry Bar in February 1992 and December 1993 (Fig. 3) and there was no decrease in P. marinus intensity. The duration of low-salinity conditions may be critical to infection intensity. The infection prevalences of P. marinus (100%) and other parasites such as Tylocephahim (17-100%) and Nematopsis (58-75%) were much higher than the less than 5% prevalences previously found at other Gulf of Mexico estuaries (Couch 1985). RGS (Fig. 5) was greatest at the initial stages of spawning (April), diminished as spawning proceeded, and recovered after spawning was completed. In 1991. gonads were enlarging in No- 552 Fisher et al. vember and December even though new gametes were not visible in the follicles until February 1992. This means that gonads de- velop long before gametogenesis or that early gametogenesis was undetectable by these histologic methods. Gonad sizes in Novem- ber and December 1992 were not as large as those in 1991, indi- cating annual variability; this has also been shown by Loosanoff (1965) using Long Island oysters. It can also be inferred from the monitoring of gonadal condition that spawning during 1992 was continuous; the depletion of the follicles was relatively constant from May through October (Fig. 6). Individual oysters may have had spawning peaks during this period, but there were few departures (e.g., August sample from Dry Bar) from the regular depletion of mature gametes. There was no indication of a second gametogenic cycle at any time during the year, even though evidence for such a possibility was previously noted for Galveston Bay oysters (Soniat and Ray 1985). The data generated by this study offer some insights to the variability of several physiologic characteristics of Apalachicola Bay oysters. Seasonal cycling in most of these characteristics can be largely attributed to effects of temperature (excluding potential differences in endogenous rhythms). Temperature may have af- fected oyster physiologic characteristics directly or indirectly through control over reproductive activities. Other environmental factors, particularly salinity, may have been involved in the dif- ferences demonstrated between sites at specific collection dates. Because not all differences could be explained by the water quality measurements of this study, other factors (particularly nutrition) must be considered to understand the variability. Successful interpretation of bivalve physiologic characteristics as indicators of their health or as biologic indicators of environ- mental contamination demands thorough knowledge of natural sources of variability in the measurements. The physiologic char- acteristics of oysters (and other bivalves) may effectively reflect anthropogenic stress, provided that variability in the indicators is either limited or well defined and correctable (Huggctt et al. 1992). Some of the characteristics measured here were signifi- cantly different at the same sampling month for different years (Table 3), likely as the result of discrepancies in temperature or other environmental conditions. These data represent only monthly samples during a year-long period for a limited geo- graphic region, so the interpretations must be verified for different years and different locations. ACKNOWLEDGMENTS We appreciate the technical, histologic, and statistical assis- tance of P. Edwards, J. Gillet, and Dr. C. Bundrick, respectively. This is EPA Gulf Ecology Division Contribution No. 972. LITERATURE CITED Andrews. J. D. & W. G. Hewatt. 1957. Oyster mortality studies in Vir- ginia II. The fungus disease caused by Dermocyslulium nmniuim In oysters of Chesapeake Bay. Ecoi. Monogr. 27:1-26. Austin, H.. D. S. Haven & M. S. Moustafa. 1993. The relationship be- tween trends in a condition index of the Amencan oyster Crassoslrea virginica, and environmental parameters in three Virginia estuaries. Estuaries 16:362-374. Barber, G. J., S. E. Ford & R. N. Wargo. 1991. Genetic vanation in the timing of gonadal maturation and spawning of the eastem oyster. Cras- soslrea virginica (Gmelm), Biol. Bull. 181:216-221. Bayne, B. L., D. R. Livingstone, M. N. Moore & J. Widdows. 1976. A cytochemical and biochemical index of stress in Mylilus edulis (L). Mar. Pollut. Bull. 7:221. Bayne, B. L., N. M. Moore & R. K. Koehn. 1981. Lysosomes and the response by Mytilus edulis L. lo an mcrease in salinity. Mar. Biol. Lett. 2:193-204. Butler, P. A. 1955. Reproductive cycle in native and transplanted oysters. Proc. Natl. Shellfish. Assoc, p. 75. Chagot, D., C. Azlieu. J. Sanjuan & H. Grizel. 1990. Sublethal and histopalhological effects of trace levels of tributyltin fluoride on adult oysters [Crassostrea gigas). Aquat. Liv. Resour. 3:121-130. Chu. F.-L. E. & K. H. Greene. 1989. Effect of temperature and sahnity on in vitro culture of the oyster pathogen, Perkinsus marinus (Api- complexa: Perkinsea). J. Invertebr. Pathol. 53:260-268. Conner, D., A. Conway, B. Benedict & B. Christensen. 1982. Modeling the Apalachicola System: a hydrodynamic and water quality model with a hydrodynamic and water quality atlas of Apalachicola Bay. Florida Sea Grant Technical Paper No. 23. 89 pp. Couch, J. A. 1985. Prospective study of infectious and noninfectious dis- eases in oysters and tlshes in three Gulf of Mexico estuanes. Dis. Aquat. Org. 1:59-82. Dawson, D. E. 1955. A contribution to the hydrography of Apalachicola Bay. Publ. Texas Inst. Mar. Sci. 4:15-35 Ednoff, M. 1984. A mariculture assessment of Apalachicola Bay, Florida. A report to the Office of Coastal Management. Flonda Department of Environmental Regulation, 239 pp. Engle, J. B. 1951. The condition of oysters as measured by the carbohy- drate cycle, the condition factor and the percent dry weight. Proc. Natl. Shellfish Assoc. 41:20-25. Estabrook, R. H. 1973. Phytoplankton ecology and hydrograph of Apalachicola Bay. M.S. Thesis, Flonda State University, Tallahassee. 166 pp. Fisher, W. S., L. M. Oliver & P. Edwards. 1996. Hematologic and se- rologic variability of eastem oysters from Apalachicola Bay, Florida. J. Shellfish Res. (In press). Florida Department of Natural Resources. 1986. Summary of Flonda com- mercial marine landings. Florida Department of Natural Resources, Tallahassee, 36 pp. Gallsoff, P. W 1964. The American Oyster Crassostrea virginica Gme- lin. U.S. Fish Wildlife Serv. Fish. Bull. 64:1-480. Gauthier, J. D. & W. S. Fisher. 1990. Hemolymph assay for diagnosis of Perkinsus marinus in oysters Crassostrea virginica (Gmelin, 1791). J. Shellfish Res. 9:367-371. Gorsllne, D. S. 1963. Oceanography of Apalachicola Bay, Florida. Oceanographic Institute Florida Stale University, Tallahassee. Con- tribution No. 196. Hopkins. S H., I. G. Mackin & R W, Menzel. 1954. The annual cycle of reproduction, growth and fattening. Proc. Natl. Shellfish Assoc. 44:39-50. Huggelt, R. J.. R. A. Kimerle. P. M Mehrle. Jr. & H. L. Bergman. 1992. Biomarkers: Biochemical, Physiological and Histological Mark- ers of Anthropogenic Stress. SETAC Special Publication Series, So- ciety of Environmental Toxicology and Chemistry, Lewis Publishers, Boca Raton, FL. Hydrolab Corporation 1991 Datasonde 3: multiparameter water quality datalogger operating manual, Hydrolab Corporation, Austin, TX, 83 pp. Ingle, R. M. & C. E. Dawson, Jr. 1952. Growth of the American oyster Crassostrea virginica In Florida waters. Bull. Mar. Sci. Gulf Carihb. 2:393-404. Ingle. R. M. & D. E, Dawson, Jr. 1953. A survey of Apalachicola Bay. Fla. State Brd. Conservation. Tech. Series 10:1-38. International Mussel Watch (IMWl, 1980. National Academy of Sciences, Washington, D.C. 248 pp. Physiologic Variability of Eastern Oysters 553 Lawrence, D. R. & G. I. Scott. 1982. The determinarion and use of con- dition index of oysters. Esluahes 5:23-27 Loosanoff, V. L. 1965. Gonad development and discharge of spawn in oysters of Long Island Sound. Biol. Bull. 129:546-561. Loosanoff. V. L. 1969. Maturation of gonads of oysters, Crassostrea virginica. of different geographical areas subjected to relatively low temperatures. Veliger 11:153-163. Lowe. J. I.. R. P. Parish. J. M. Patrick & J. Forester, 1972. Effects of polychlorinated biphenyl Arochlor 1254 on the American oyster. Cnis- soslrea virginica. Mar. Biol. 17:209-214. Paynter. K. T. & E. M. Burreson. 1991. Effects of Perkinsiis marinus infection in the eastern oyster. Crassostrea virginica: II. Disease de- velopment and impact on growth rate at different salinities. J. Shellfish Res. 10:425^31. Perkins. F. O. 1966. Life history studies of Dennocystidnim marinum. an oyster pathogen. Ph.D. Dissertation. Flonda State University. Talla- hassee, FL. Quick, J. A., Jr. 1971. A Preliminary Investigation: The Effect of Ele- vated Temperature on the American Oyster Crassostrea virginica Gmelin. Florida Dep Nat Res. Professional Papers Series. No. 15. 190 pp. Ray, S. M. 1966. A review of the culture method for detecting Dermo- cystidium marinum. with suggested modifications and precautions. Proc. Natl. Shellfish. Assoc. 54:55-69. Rockwood. C. E. & S. Leitman. 1977. Economic planning for the Apalachicola drainage system. In: R. J. Livingston & E. A. Joyce, Jr. (eds.). Proceedings of the Conference on the Apalachicola Drainage System Flonda Mar. Res. Publ. No. 26. pp. 151-157. Roper, D. S.. R. D, Pndmore, V. J. Cummings & J. E. Hewitt. 1991. Pollution related differences in the condition cycles of Pacific oysters Crassostrea gigas from Manuhau Harbor, New Zealand. Mar. Env. Res. 31:197-214. Ruddy, G. M., S. Y. Feng & G. S. Campbell. 1975. The effect of pro- longed exposure to elevated temperatures on the biochemical constit- uents, gonadal development and shell deposition of the American oys- ter, Crassostrea virginica. Comp. Biochem. Physiol. 518:157-164. Scott. G. I. & D. P. Middaugh. 1978. Seasonal chronic toxicity of chlo- rination to the American oyster Crassostrea virginica. pp. 311-328. In: Water Chlorination. vol. 2. Environmental Impact and Health Ef- fects. Ann Arbor. MI. Scott. G. I. & W. B. Vemberg. 1979. Co-occuring chlorine produced oxidants in seawater and their effect on growth, survival, and physi- ology of the .'\niencan oyster. Crassostrea virginica (Gmelin). pp. 415^35. In: Marine Pollution: Functional Responses. Academic Press, New York. Soniat, T. M. & S. M. Ray. 1985. Relationships between possible avail- able food and the composition, condition and reproductive state of oysters from Galveston Bay, Texas. Contrib. Mar. Sci. 28:109-121. Thompson. R.. V Brooks. J. Gunter & E. Baraett. 1990. Comprehensive Shellfish Harvesting Area Survey. Apalachicola Bay. Franklin County. Flonda Department of Natural Resources, Shellfish Environmental As- sessment Section. Tallahassee. FL. Thompson. R. J.. N. A. Ratcliffe & B. A. Bayne. 1974. Effects of star- vation on structure and function in the digestive gland of the mussel {Mytdus ediilis L.). J. Mar. Biol. Assoc. U.K. 54:699-712. U.S. Army Corps of Engineers. 1978. Coordination report on navigational improvements for Apalachicola River below Jim Woodruff Dam. Flor- ida. U.S. Army Corps of Engineers, Mobile District. Weis, P., J. S. Weis, J. Couch, C. Daniels & T. Chen. 1995. Pathological and genotoxological observations in oysters {Crassostrea virgirtica) living on chromated copper arsenate (CCA)-lreated wood. Mar. Env. Res. 39:275-278. Whitfield. W. K., Jr. & D. S. Beaumariage. 1977. Shellfish management in Apalachicola Bay: past, present, future, pp. 130-140. In: R. J. Livingston & E. A. Joyce, Jr. (eds.). Proceedings of the Conference on the Apalachicola Drainage System. Florida Marine Research Pub- lications No. 26. Widdows. J. & D. Johnson. 1988. Physiological energetics of Mytilus ediilis: Scope for growth. Mar. Ecol. Prog. Ser. 46:11.3-121. Winstead, J, T. 1995. Digestive tubule atrophy in eastern oysters, Cras- sostrea virginica (Gmelin, 1791) exposed to salinity and starvation stress. J. Shellfish Res. 14:105-111. Journal of Slwllfish Rcsecinh. Viil. 15, No. 3. 555-564, 1996. HEMATOLOGIC AND SEROLOGIC VARIABILITY OF EASTERN OYSTERS FROM APALACHICOLA BAY, FLORIDA WILLIAM S. FISHER,' LEAH M. OLIVER,' AND PATRICE EDWARDS- ' U.S. Envirowneniul Protection Agency National Health and Environmental Ejfects Research Laboratory Gulf Ecology Division Gulf Breeze. Florida 32561 ^Florida Marine Research Institute Department of Environmental Protection Ft. Walton Beach. Florida 32548 .ABSTRACT Eastern oysters iCrassoslrea virginica) were collected monthly from two sites approximately 15 km apart in Apalach- icola Bay, FL, during a 1-y period. Hematologic and serologic measurements were made on hemolymph withdrawn from the adductor muscle. The two sites experienced nearly identical temperature patterns during the study penod, but salinity and other physical factors fluctuated. Significant differences attributable to sampling date were found for circulating hemocyte density, phagocytic activity, and superoxide anion (O, )-producing ability and for serum protein, lysozyme. and agglutinating activity, with data from both sites combined. This variability was most likely related to temperature or temperature-influenced reproductive cycling. Oyster hemocyte locomotion did not vary significantly with lime over the study period, nor were significant differences found between sites. Significant differences between site means (combined for all dates) were found for O,", protein and lysozyme, and significant date * site interactions were found for phagocytosis, agglutination, and lysozyme, indicating that local conditions, such as salinity lluctuations, influenced these measurements. An accurate description of variability in oyster defensive functions will require more frequent sampling and a better understanding of local environmental intluences. KEY WORDS: Eastern oysters {Crassosrrea virginica). bivalve immunology, invertebrate hematology, invertebrate serology INTRODUCTION Several hematologic and serologic parameters have been mea- sured to indicate the status of the defense (immune) system of bivalve molluscs (for reviews, see Chu 1988 and Feng 1988). Bivalve defensive cells and hemolymph molecules are influenced by changes in ambient temperature and salinity (Fisher 1988). which may account for the high variability in the defense re- sponses of oysters collected at different sites and/or times of year (Fisher et al. 1989, Oliver and Fisher 1995). This variability is a critical concern for research that attempts to understand the rela- tionship of defense processes with infection and resistance in oys- ters (Crassostreii virginica) afflicted by protozoan parasite dis- eases (Ford 1986. Ford 1988. Chu and La Peyre 1989, Chu and La Peyre 1993a, Chu and La Peyre 1993b. Kanaley and Ford 1990. Chintala and Fisher 1991. Anderson et al. 1992b, Chu et al. 1993. Ford et al. 1993. Chintala et al. 1994). Also, there is continued interest in descnbing the effects of xenobiotic exposure on bivalve defense mechanisms or in using measurements of defense activi- ties to characterize environmental conditions (Ruddell and Rains 1975. Fries and Tripp 1980. Anderson et al. 1981. McCormick- Ray 1987. Cheng 1988. Cheng 1990, Cheng 1993. Seller and Morse 1988. Suresh and Mohandas 1990. Sami et al. 1992; afso see reviews by Anderson 1988 and Anderson 1993). Such appli- cations of defense characteristics demand a better understanding of their ranges and sources of variability. A major source of variability in bivalve biology is the annual reproductive cycle, driven primarily by temperature (Galtsoff 1964). Many physiologic and defense characteristics are influ- enced by this dominating activity (Eble 1966, Swift and Ahmed 1983. Fisher and Newell 1986a. Fisher et al. 1989). but intermit- tent changes in other environmental factors such as salinity. dissolved oxygen, nutrients, toxicants, parasites, and disease also exert .some influence. The combined effects of these exoge- nous factors modify the inherent defense capabilities of each or- ganism. Variability in defense characteristics has been demonstrated in vitro after various alterations in salinity and temperature condi- tions (Fisher and Newell 1986b. Fisher 1988. Fisher and Tamplin 1988). Several studies have shown variable hemocyte responses of oysters collected at several locations or held in the laboratory under different controlled conditions (Fisher and Newell 1986a, Fisher et al. 1989. Chu and La Peyre 1993a. Chu et al. 1993, Cheng et al. 1993, Oliver and Fisher 1995). Yet, for estuarine oysters that can experience diurnal tidal fluctuations, frequent and sometimes drastic alterations in nature make it difficult to extrap- olate laboratory findings to field conditions. Oysters from two bars in Apalachicola Bay (Franklin County, FL) were studied from October 1991 to March 1993 to examine variations in hemocyte and hemolymph characteristics. The oyster bars lie within 15 km of each other in a bay system protected by a chain of barrier islands ( Gorsline 1 963 ) . Temperature differences between these sites during the collection period were minimal, and oyster reproductive cycles were nearly identical (Fisher et al. 1996). However, currents, tides and river inflow created smaller- scale and intermittent environmental diversity in other factors such as salinity. The goals of this research were to monitor and describe variability in oyster defense activities and to distinguish contribu- tions of seasonal temperature cycles from site-specific factors. Presented here are monthly averages for circulating hemocyte number, locomotive activity, phagocytic activity, and superoxide anion (Oj ) generation, as well as hemolymph protein concen- tration, lysozyme concentration, and agglutinin titer to horse erythrocytes. 555 556 Fisher et al. MATERIALS AND METHODS Background Information The Apalachicola Bay system is a highly productive lagoon/ barrier island complex. The eastern oyster is the most important commercial invertebrate in Apalachicola Bay and comprises nearly 90% of the oysters harvested in Florida. Because of rela- tively mild temperatures in the area, rapid oyster growth is sus- tained throughout the year. Harvestable oysters, those larger than 3 inches (76 mm), have been grown from spat in as little as 39 wk, and the spawning season is reportedly one of the longest in the United States (Ingle and Dawson 1952). Oysters in Apalachicola Bay. like most oysters throughout the Gulf of Mexico, exhibit a nearly 100% prevalence of the protozoan pathogen Perkinsus imiriiuis (Craig et al. 1989). Site Characteristics The two sites for oyster collection. Cat Point Bar and Dry Bar (also called St. Vincent's Bar), were previously described in re- lation to hydrology, water quality, and value to the oyster industry (Fisher et al. 1996). A history of water quality data exists through the efforts of the Florida Department of Environmental Protection (Thompson et al. 1990). The two sites are less than 15 km apart and are southeast (Cat Point Bar) and southwest (Dry Bar) of the mouth of the Apalachicola River (Fig. 1 ). Both sites lie within the chain of protective barrier islands, but the physical and chemical characteristics of the water are influenced by wind, tide, and river currents (Gorsline 1963). For this study, temperature and salinity were measured at the time of oyster collection. Also, continuous (30-min interval) monitoring of water quality was initiated in June 1992 and continued through December 1992 with Hydrolab' Data- sonde 3 dataloggers attached to pilings at the collection sites (Fisher etal. 1996). Oyster Collection and Processing Oysters were collected with hand tongs on (approximately) a monthly basis from October 1991 to October 1992. and additional samples were collected in December 1992 and March 1993. Dates of collection were (1991): October 22. November 19, December 19;(1992): January 21, February 25, March 31, April 21. May 19, June 23. July 21. September 1. September 29. October 26, De- cember 7; and ( 1993): March 29, for a total of 15 collections over a span of 17 mo (525 days). The September I sample is referred to as the August sample in this report. Oysters ranging from 50 to 90 mm in height (umbo to beak) were collected and placed im- mediately into coolers containing cold ice packs. The coolers were transported to the Environmental Protection Agency Gulf Ecology Division at Gulf Breeze, FL, and placed in a refrigerator (4°C) overnight. Twelve oysters from each site were allowed to warm to room 'The mention of product names does nol signify a recommendation or endorsemenl hy the Environmental Protection Agency. Figure 1. The Apalachicola River and Bay system with the two oyster collection sites. Cat Point Bar and Dry Bar (also known as St. Vincent's Bar), noted with stars. The sites are in northwest Florida (inset). Variability of Eastern Oysters 557 temperature for 2^ h. They were then scrubbed clean of fouHng organisms, and a grinder was used to notch the shells at the pos- terior edge adjacent to the adductor muscle. The mantle cavity was rinsed thoroughly with filtered (0.22 (Jtm pore size) seawater to remove debris. Hemolymph was withdrawn from the adductor muscle with a 3-mL syringe and a 22-gauge needle. Hemolymph was used to measure vanous hematologic and serologic character- istics, as described below, and to diagnose P. iminnus infection intensity (see Fisher et al. 1996). These characteristics were mea- sured for different durations during the course of the study (see Table 1). Circulating Hemocyle Density Hemolymph withdrawn from oysters was mixed gently to en- sure sample homogeneity; then, one drop was dispensed onto each side of a hemacytometer counting chamber for duplicate counts. The remaining hemolymph sample was placed on ice immediately after hemacytometer loading to minimize hemocyte aggregation and activity. Hemocyle Mobility and Rale of Location Two drops of hemolymph were placed in a single well of a Lab-Tek 8 chamber slide containing 200 |xL of filtered (0.45 (xm pore size) water collected from the site. Hemocytes were main tained at 27°C for 30-60 min to allow hemocyte settling, attach- ment, and locomotion along the glass slide surface. For each oys- ter, the movement of 12 to 15 adherent hemocytes was tracked with transparent overlays on a video monitor attached to an Nikon inverted microscope under phase contrast (Fisher and Newell 1986b). Tracings were measured with electronic digital calipers. The average rate for at least 10 mobile hemocytes was recorded as the rate for each individual oyster. Hemocytes that did not move were also recorded, and the percentage of mobile hemocytes was calculated. Particle-Binding (Phagocytic) Index The ability to bind a foreign particle (phagocytic index. PI) was determined in vitro by the addition of a calculated volume of hemolymph containing 30,000 hemocytes to wells of Lab-Tek 8 chamber slides that had been preloaded with 30,000 test particles (yeast, Succharomyces cerevisiae; Sigma Chemical Co.) sus- pended in filtered seawater sampled from the collection site (see Oliver and Fisher 1995). The proportion of [yeast:hemocyte] was maintained at |l:l) because phagocytic activity is dependent on particle availability. After 60 min of incubation at 27°C, the slides were gently dipped in filtered seawater to remove unbound parti- cles and nonadherent hemocytes, fixed for 2 min in absolute meth- anol, and then dipped in deionized water and fresh methanol be- fore air drying. Examination of slides was performed on a Nikon inverted microscope with 40 x objective under phase contrast. A minimum of 200 hemocytes was examined per well to determine the PI, which was expressed as: (number of hemocytes displaying bound or ingested yeast) x 100 (total number of hemocytes observed) Hemocyte Superoxide Anion Production Superoxide anion (O,") generation by hemocytes from indi- vidual oysters was quantified by measuring the reduction of ni- troblue tetrazolium INBT) dye to formazan. The colorimetric method described by Anderson et al. (1992a) was scaled for an analysis of individual oysters as described by Oliver and Fisher (1995). Hemocyte production of O, " was measured under both unchallenged and yeast-challenged conditions. Hemolymph Serum Protein Hemolymph (0.25-0.5 mL) from each individual was placed in a plastic microfuge tube and centrifuged at 2,900 x ^ for 5 min. The resulting pellet was cultured in Ray's fluid thioglycollate me- dium (Ray 1966) to diagnose the intensity off. marimis disease (Fisher et al. 1992), and the serum (supemate) was held at 4°C for less than 1 wk before being tested for protein, lysozyme, and agglutinin content. Levels of serum protein were measured with the Pierce BCA Protein Assay kit with samples diluted 10- fold in deionized water. Protein concentrations for each sample were cal- culated from a standard curve generated from dilutions of bovine serum albumin. TABLE 1. Results of an interactive ANOVA with date and site for the variables HC, percent mobile hemocytes (% mobile), rate of hemocyte locomotion (RHL), PI, reduction of NBT, unchallenged (NBT-UNCH) and challenged with yeast (NBT-CHALL), hemolymph serum protein content, agglutinin content, and lysozyme content." Variable Duration Date (p<) Site (p<) Date * Site (p<) 0.167 NS' 0.858 NS^ 0.104 NS' 0.015'' 0.082 NS' 0.171 NS' 0.259 NS' 0.001" 0.002" HC % Mobile RHL PI NBT-UNCH NBT-CHALL Protein Agglutinin Lysozyme Dec '91 Dec '91- Dec '91- Jan '92- Mar '92- Mar '92- Oct '91- Oct '91- Apr '92- -Mar '93 -Mar '93 -Mar '93 Mar '93 -Mar '93 -Mar '93 Mar '93 Oct '92 -Mar '93 0.001" 0.182 NS' 0.161 NS' 0.001" 0.001" 0.001" 0.001" 0.001" 0.001" 0.916 NS" 0.674 NS" 0.636 NS' 0.224 NS' 0.007" 0.021'' 0.001" 0.338 NS' 0.011'' ■■ Tests were performed for different durations over the study period. " Highly significant (p < 0.01). ■" NS, not significant. ' Signitlcani (p < 0.05). 558 Fisher et al. Agglutination by Hemolymph Serum Hemolymph serum samples were tested for the agglutination of horse er>'throcytes (red blood cells [RBC]) obtained from Cocalico Biologicals. Inc. (Reamstown, PA) and diluted 1:1 in Alsever's solution. Serial twofold dilutions of 50 yiL of serum in 96-well microtiter plates were assayed with 50 (xL of a 2% RBC suspen- sion in 8.5 ppt artificial seawater, according to previously pub- lished methods (Fisher and DiNuzzo 1991, Fisher 1992). Micro- titer plates were examined after 2-3 h of incubation at 24°C, and titers were recorded as the reciprocal of the highest dilution show- ing positive agglutination. Titers were recorded as log, values to reflect the twofold serial dilutions. Hemolymph Serum Lysozyme Lysozyme was quantified by measuring the ability of oyster serum to degrade a suspension of bacteria Micrococcus lysodeik- ticiis (Sigma Chemical Co.). These procedures were originally described by Shugar (1952) and modified for oyster serum by Rodrick and Cheng (1974). Hemolymph serum samples were stored in the refrigerator for 1-2 days before quantification. A 0.2 mg/mL bacterial suspension was prepared by adding 0.005 g M. lysodeiklicus to 25 mL of 0.5% glycylglycine buffer (Gly) (Gly = 1.65 g of glycylglycine [Aldnch Chemical Co.| plus 0.625 g of sodium chloride in 125 mL of deionized water). The pH of Gly was adjusted to 5.5 with IN hydrochloric acid. The bacteria were resuspended in Gly, resulting in a uniformly turbid solution, and the optical density was adjusted to 0.7 absorbance units at 540 nm with a Shimadzu spectrophotometer. A disposable microcuvette was loaded with 20 |xL of serum, followed by 0.5 mL of M. lysodeiklicus. The decrease in turbidity in the microcuvette was traced for 4 min at 540 nm in the kinetic mode of the spectropho- tometer. The rates of degradation were compared with those of standard solutions prepared from hen egg white lysozyme (Sigma Chemical Co.). Statistical Methods Data were entered into SAS (SAS Institute. Cary. NC) and analyzed with General Linear Models and the Shapiro-Wilk test for normalcy. Residual plots were examined to check for homo- geneity of variance, independence, and normality of error terms in the resulting models. Two-way analysis of variance (ANOVA) was conducted to relate each dependent variable (hemocyte and hemolymph measurements) to the main effects date and site and to test for possible interactions between date and site ( =date * site). The results of all analyses are reported, but main effects are dis- cussed only if there was no significant interaction. Where signif- icant main effects were found, Turkey's post-hoc test was used to differentiate between overall date or site means. For variables that displayed a significant interaction effect, differences due to date for each of the sites were examined with Tukey's post-hoc results, which compared means of all date * site combinations. Levels of significance and high significance were designated as p =s 0.05 and p =£ 0.01, respectively. RESULTS Temperature and Salinity Conditions Water temperatures measured when oysters were collected were very consistent for the two collection sites (Fig. 2. top), and close agreement in daily values was confirmed by Hydrolab data for part of the study period (Fig. 2. bottom). Both sites experi- enced progressive warming after January and cooling after July. In contrast, salinity measured at the time of collection varied greatly between the two sites (Fig. 3. top); data collected continuously from June 1992 to March 1993 revealed differences between the two sites' daily averaged salinity values that were often 10-15 ppt and sometimes as high as 25 ppt (Fig. 3. bottom). Even within a given day. 10-15 ppt salinity variation at one site was not uncom- mon. Summary of Oyster Physiology Measurements of oyster physiology (see Fisher et al. 1996) showed that oysters collected from both sites were reproductively similar; gametogenesis was first visible in histologic sections in February 1992. and spawning was apparently continuous from April through September. Oysters collected were predominantly female (80%) throughout most of the study period, even though the gender ratio was approximately equal during February and March. The condition index of oysters was lowest during autumn, and the wet:dry tissue weight ratio (used to calculate total tissue weight for the condition index) generally reflected the inverse of condition index. Circulating Hemocyte Density The effect of collection date on circulating hemocyte density (HC) was highly significant, whereas no significant effect due to site or date * site interaction was found (Table I). Temporal changes were very similar for the HC of oysters from both sites (Fig. 4).^ For both sites, the highest HC was measured in May, after which it declined through the summer to the lowest level, measured in August (Table 2A); the level then increased in Sep- tember and October. Hemocyte Locomotion No significant effects were found for percent mobile hemocytes (Fig. 5) or hemocyte rate of locomotion (Fig. 6) as the result of date. site, or date * site interaction (Table 1). Particle-Binding (Phagocytic) Index A 1 1 : 1 ] yeast: hemocyte challenge resulted in 20 to 35% phago- cytic hemocytes throughout the study (Fig. 7). The effect of col- lection date on PI was highly significant, and the interactive effect of date * site was also significant (Table 1 ). Hemocytes from both sites showed a significant decrease in particle binding between April and June and recovery by August (Table 2B). Hemocyte Superoxide Anion Production Unchallenged and yeast-challenged hemocytes produced super- oxide anion (0-," ) at nearly the same levels throughout the study period (Fig. 8). Levels from challenged hemocytes were signifi- cantly correlated with those from unchallenged hemocytes (r - 0.861. n = 224. Pearson's procedure). For unchallenged and challenged 0,~ production, the main effects of both date and site were significant and no significant interactive date * site effect was found (Table 1). The overall seasonal pattern was similar for both sites, with progressively lower measurements obtained from "Data are presenled in graphic form for convenience; tabular data sets are available upon request to the authors. Variability of Eastern Oysters 559 o E M J J Date -■- Cat Point Bar -- DryBar Dry Bar 1 ■/!■ J : !■ ': ,; V Figure 2. Water temperature of Cat Point Bar and Dry Bar during the period of collection. (Top) Temperature taken at the time of col- lection throughout the entire study period. (Bottom) Daily averages of temperatures taken every 30 min from June 1992 to March 1993 only (note scale differences). All temperatures were recorded less than 0.5 m from the bay bottom. spring to summer, following by increased activity during the fall and winter months (Table 2A). The mean response over all dates for both unchallenged and challenged O, " production was signif- icantly higher for Dry Bar oysters. Serum Protein Concentration Highly significant differences in protein content were attribut- able to both date and site, and no significant date * site interaction was found (Table 1 ). Average protein content throughout the study period was significantly higher for Dry Bar oysters. With the exception of relatively low March 1992 results, the highest protein concentrations were measured during winter and spring (Decem- ber to April 1992) and the lowest were measured during the late summer (Table 2A; Fig. 9). Serum Agglutination Activity Differences in the ability of oyster sera to agglutinate horse erythrocytes were significant for date effect and date * site inter- action, but were not significant for site effect (Table I ). Although a trend appears consistent at both sites (Fig. 10). there were Cat Point Dry Bar Figure 3. Water salinity of Cat Point Bar and Dry Bar during the period of collection. (Top) Salinities measured at the time of collection throughout the study period. (Bottom) Daily averages of salinities measured every 30 min from June 1992 to March 1993 only (note scale differences). .All salinities were recorded less than 0.5 m from the bottom of the bav. few significant differences at either site over time when all date * site combinations were considered (Table 2B). Serum Lysozyme Concentration The effects of collection date. site, and date * site interaction on oyster serum lysozyme concentration variability were all sig- nificant (Table 1). Lysozyme levels increased significantly from May to July for Cat Point Bar oysters, and although a similar trend occurred for Dry Bar oysters (Fig. II). it was not significant (Table 2B). DISCUSSION Natural variations due to season and habitat cannot be over- looked if measurements of oyster defense mechanisms are to be properly interpreted in the context of physiology or disease sus- ceptibility or as biologic indicators of pollution. To achieve some insight to the natural variation of oyster hematologic and serologic characteristics, oysters from two relatively pristine sites in Apalachicola Bay were examined monthly during a 1-y period. This period encompassed the oyster reproductive cycle, which is strongly regulated by seasonal temperatures. Most of the measured defense traits varied significantly throughout the year (Table 1. 560 Fisher et al. // ^ 3.5 Cat Point Bar — ■— Dry Bar --*-- ^ 3.0 - o X 2.5 - i 7 \ T cc T I A \t S 2.0- -^ t^ \\ ^ 2 '>. \] [^^ v \ ynk \ , I——" 1 ^ 1.5 J V r^ ^ H r* i k^ LU * ' ' YV ^ o ■-^ 9 0.5 - ^^ 2 o LU z // -^ 0.0 1 1 1 1 1 1 i 1 1 // 1 '^'^^'o^W Figure 4. Circulating HC in hemolymph of oysters collected from two sites. Cat Point Bar and Dry Bar, during the period from December 1991 through March 1993. ND, no data; bars indicate standard er- rors. variability by date), and some of these exhibited apparent annual patterns that may reflect changes in seasonal temperatures or oys- ter reproductive condition. Water temperature and reproductive condition were nearly identical for the two sites during the study period (Fisher et al. 1996). However, significant site effects and interactive date * site effects (Table 1 ) demonstrated the influence of unique local conditions, which could be any combination ot physical, chemical, and biologic factors that occurred indepen- dently at each site. The relationship of salinity with site effects could not be determined in this study, but it is possible that chang- ing estuarinc salinities were partly responsible for the site effects because salinity fluctuated widely between sites (Fig. 3) and is known to have an explicit effect on defense mechanisms in vitro (Fisher and Newell 1986b. Fisher 1988. Fisher and Tamplin 1988, Fisher et al. 1989). Alternatively, salinity fluctuations may have tracked other hydrographic changes that affected oyster biology (e.g., changes in turbidity, oxygen, nutrient availability and com- position, contaminants, or other factors that might be associated with changes in river flow or currents). Variation in oyster HC and type has been demonstrated in other studies: Chu and La Peyre (1993b) found differences among oys- ters at three sites in Chesapeake Bay, and Oliver and Fisher (1995), as an adjunct to this project, found differences between hemocyte morphology and activities in oysters from Chesapeake Bay and Apalachicola Bay (Cat Point oysters) in March and Oc- tober 1992. In this study, HC was lowest in July and August (Fig. 4). coinciding with the highest water temperatures during the year and active spawning by the oysters. Higher HC recorded in Sep- tember could have been related to decreasing temperature and the end of spawning, when they resorb gonadal tissue. The lack of significant site or date * site effects implies that variation in HC was associated more with seasonal temperatures than with local conditions. This pattern contrasts with the relatively high HC found for oysters held in warm temperatures by Chu and La Peyre ( 1993b) and with earlier observations made by Feng ( 1965), who related increased temperature to a higher oyster heart rate, which propelled a larger number of hemocytes into circulation. Because those studies used oysters from the mid- and north Atlantic, it is possible that the contrasting results are due to higher mean summer temperatures in the Gulf of Mexico or to the unique physiologic characteristics of each oyster population. Hemocyte mobility (percent mobile hemocytes and rate of lo- comotion) had no significant variability attributable to date or site. Earlier work (Fisher et al. 1989) with oyster hemocyte samples from an oceanic and an estuarine environment also found no effect on the rate of locomotion by warm summer temperatures, even though hemocyte-spreading capacity was retarded. Acute in vitro increases in salinity have been shown to depress hemocyte loco- motion in laboratory studies (Fisher and Newell 1986a, Fisher and Tamplin 1988. Fisher et al. 1989). but it is not likely that such effects could be detected with a monthly monitoring regimen. The phagocytic ability (PI) of oyster hemocytes varied signif- icantly by date, but a clear seasonal pattern was not evident (Fig. 7). Phagocytic activity by Dry Bar oyster hemocytes was signifi- cantly lower in March 1993 compared with March 1992 (Table 2B). demonstrating the likelihood of year-to-year variability. The significant date * site interaction (Table 1) indicates that local conditions exerted an inconsistent influence on PI measurements. The challenged and unchallenged production of O," by hemo- cytes, measurements to reflect phagocytosis-associated killing ca- pacity, demonstrated an apparent seasonal pattern (Fig. 8), as well as a significant site difference (Table 1). It appears that O," 70 en 60 Ul ^ 50 u ^ 40 m X LU .30 _J m O 2U 10 Cat Point Bar Dry Bar —^ -//- 4^i^Mif^^i<^M Figure 5. Percentage of circulating hemocytes that exhibited mobility. Hemocytes were withdrawn from oysters collected from two sites. Cat Point Bar and Dry Bar, during the period from December 1991 through March 1993. ND, no data; bars indicate standard errors. 7 6 5 LU ^ 4 z 2 3 Cat Point Bar Dry Bar — ^- -YA- ~i — r -Yy^ o^VA^/Z/^^V'^o^'/^ ^^ Figure 6. Rate of locomotion of circulating hemocytes drawn from oysters collected from two sites. Cat Point Bar and Dry Bar. during the period from December 1991 through March 1993. ND, no data; bars indicate standard errors. Variability of Eastern Oysters 561 production by oyster hemocytes was closely related to temperature because significant increases occurred in August and September, just as temperatures began to decline. Oysters at both sites were still spawning through September, so the effect cannot be clearly linked to reproductive cycling. In contrast. Anderson et al. (1992a) reported that the O, production of hemocytes from Chesapeake Bay oysters, measured by NBT reduction of hemo- cytes pooled from several oysters, was significantly greater in those samples collected from waters at 2I-29°C as compared with those at 2-1 3°C. Again, because those studies used oysters from the mid- Atlantic, it is possible that the contrasting results are due to higher mean summer temperatures in the Gulf of Mexico or to the unique physiologic characteristics of each oyster population. For example, Oliver and Fisher (1995) found O," production by Chesapeake Bay oyster hemocytes to be twice that of Apalachicola Bay oyster hemocytes in both March and October 1992. Volety and Chu (1995) presented evidence that the oyster pathogen P. inahnus could suppress the production of reactive oxygen intermediates (including O, ^ ) in vitro, even though hemo- cytes taken from oysters with heavy P. marmus infections were found to have increased O, " production compared w ith individ- uals with light infections (Anderson et al. 1995). In this study, the prevalence and intensity off. marmus (measured by hemolymph assay) changed significantly over time (Fisher et al. 1996). but there was no obvious trend or pattern that might be related to the changes in O, " production observed. Obviously, the factors con- tributing to variability in this important defense activity must be further elucidated. Serum protein concentrations appeared to exhibit an annual pattern for oysters at both sites, with the highest levels recorded in February and the lowest recorded m August (Table 2). A similar phenomenon was described in a study of Chesapeake Bay oysters (Fisher and Newell 1986a) that suggested that the high levels were related to the final stages of gamete maturation. Chu and La Peyre (1989) also found peak serum protein concentrations in Chesa- peake Bay oysters from February to March, with lower values in June and July. It is not known why protein concentrations were higher in serum from Dry Bar oysters. Site-specific factors, evidenced by a significant date * site in- teraction (Table I), may have contributed to the variability ob- served in the agglutination of horse RBC by oyster serum (Fig. CO m o O o o o o < X CL bU -n Cat Point Bar — ■— Dry Bar --^— 40 - 30 - J f*wA A^ A 20 - ^\f%*^^ 10 - 0 - Q Z 1 Q I 1 1 1 1 1 1 1 1 1 /X 1 o .<^v ^^.<^*v^^^V.O^^4 "^ 0.04 - J K. Vr^^ -^ I 0.03 -\ r^b^ r 0.02 - ^^ 1 0.01 - Q Q z: Q Z Q Z // 1 1 1 1 1 1 1 1 1 1 // 1 \<^'o^VV.'^V//#.^>%^'o^V# Figure 8. Spectrophotometer absorbance readings representing the amount of NBT reduced (O, produced) by oyster hemocytes. (Top) Unchallenged hemocytes. (Bottom) Hemocytes challenged with yeast. Hemocytes were drawn from oysters collected at two sites. Cat Point Bar and Dry Bar. during the period from March 1992 through March 1993. ND, no data; bars indicate standard errors. 10). Agglutinating molecules are believed to function in bivalve defense by opsonizing foreign particles for phagocytosis (Tripp 1966, Tripp and Kent 1967, Anderson and Good 1976, Olafsen 1988, Olafsen et al. 1992). On the basis of correlational observa- tions, such a role was hypothesized for C virginica as a defense against disease caused by Haplosporidium nelsoni (Chintala and Fisher 1991); however, this could not be confirmed in field studies (Chintala et al. 1994). Agglutinins may also be secreted by epi- thelial mucous cells to selectivity bind nutrient particles for inges- tion (Fisher 1992). If so. then, nutrient composition and availabil- ity at different sites could have played an important role in the variability observed. Delaware Bay oysters (Feng and Canzonier 1970) and Chesa- peake Bay oysters (Chu and La Peyre 1989) have exhibited very low. sometimes undetectable concentrations of hemolymph lysozyme during summer months. Similarly, Chu and La Peyre (1993a) have demonstrated in the laboratory, using Chesapeake Bay oysters, that serum lysozyme has an inverse relationship with temperature. In contrast, results for Apalachicola Bay oysters high lysozyme levels during the warmest month (July) and lower levels during fall and winter (Fig. II). High lysozyme levels during summer could be explained by increased secretion from hemocytes due to P. marinus invasion. Alternatively, high levels could rep- resent a breakdown of lysosomes due to stress (decreased lyso- 562 Fisher et al. TABLE 2. Study results." A PI B AGGL LYSO Date HC NBT-UN NBT-CH PROT CP DB CP DB CP DB Oct '91 D ABC C Nov CD ABC AB Dec AB ABCD ABC A Jan '92 AB AB ABCDEFG ABCDE ABC ABC Feb ABC A ABC A BC ABC Mar AB AB ABC BCD ABCDEFG BCDEFGH BC ABC Apr AB AB ABC ABC AB ABCDEF ABC ABC CDE BCDE May A BC BCD CD CDEFGHI BCDEFGHI ABC AB CDE BCDE Jun ABC C CD BCD GHI HI ABC ABC ABC BCD Jul BC C D ABCD GHI BCDEFGH AB ABC A AB Aug C ABC ABC D ABCDE ABCD ABC A BCDE ABC Sep ABC AB A CD EFGHI ABCDEFG AB ABC CDE CDE Oct AB A A BCD DEFGHI CDEFGHI A AB CDE ABC Dec ABC A A ABCD CDEFGHI ABCDEFG E DE Mar '03 BC A AB BCD FGHI I BCDE ABCD " (A) Significant differences due to sampling date (Tukey's procedure). Shared letters in the column indicate no significant differences between collection dates for that variable. Analysis was performed with site data combined for variables with no significant date * site interaction (see Table 1). (B) Significant differences between means from all date-site combinations with significant mteraction (see Table I). Shared letters between any date-site combinations mdicate no significanl differences. NBT-UN, reduction of NBT. unchallenged; NBT-CH, reduction of NBT, challenged with yeast; PROT, protein concentration; AGGL, agglutination of horse erythrocytes; LYSO, lysozyme concentration; CP, Cat Point Bar oysters; DB, Dry Bar oysters. somal latency) or increased secretion associated with feeding and spawning (Eble 1966). The variability described here and the strik- ing contrast with studies of mid-Atlantic oysters present intriguing questions related to the source and/or function of serum lysozynies. Much of the variability described in these defense-related mea- surements could stem from changes in hcmocyte composition rather than actual changes in hcmocyte activity. For example, it is generally considered that granular hemocytes are more phagocytic than agranular hemocytes (Renwrantz et al. 1979) and generate more O," (Anderson et al. 1992a). There is some evidence that circulating hemocyte composition changes over time: McCormick- Ray and Howard (1991) found that the percent granulocytes in Chesapeake Bay oyster hemolymph decreased from January to May. Also. Oliver and Fisher (1995) demonstrated a disparity between the hemocyte composition of oysters from Chesapeake Bay and Apalachicola Bay. as well as a change in the composition at both sites between March and October. However, identifying changes in hemocyte composition is complex because of polymor- phic cells and varying definitions for morphological observations, including "granularity" (M. Auffret, Universite Bretagne Occi- dentale. pers. comm.). For Gulf of Mexico oysters, the capacity to identify cell types is more difficult because hemocytes generally have less cytoplasmic volume with few morphological distinc- tions. A major role for oyster hemocytes is to respond defensively to 10.0 - Gat Po int Bar -^ Dry Bar ^*— -^^ 7.5 - } r 'i b ^ x^ \ 5.0 - V K Y^ \^ '\^ \'r H 2.5 - r* T . . . I 0.0 - 1 1 1 1 1 1 1 1 1 1 o z 1 1 r^^^ o DC CL E E Figure 9. Protein concentrations in hemolymph of oysters collected at two sites. Cat Point Bar and Dry Bar, during the period from October 1991 through March 1993. ND, no data; bars indicate standard er- rors. 9 n Gal roinl Bai ■ Uty Da( '^ ~ 8 - z T T O 7 - k 1 r y ■^ T J 1 1- \A ^V- < 6 - \/ \ r\f i z r • \ 7 ^ 1 1/ )L^ ^ 5- /f \ T / ~i/ V /T " /r\ \ T " /j h'\ i I _i ' / \ \ l^ V A (5 4- / \ ff T . A 1 ■- ) \7 O r jL > ' / w < 3 - 1 ^ ^ /J L i ^ o / ' '\ 1 i 1 O 2 - / J L \ y _i J ' 1 1 - 1 Q Q Q t z z ^^z 0 - 1 1 1 1 1 1 1 1 1 1 1 1 1 1 y;^^^ o fV>V.<^VT/#.^>%<^V>V o I \ \ 1 1 1 \ 1 1 — 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 year Figure 5. (a) Mean spatfall, York River, VA, 1946-1992, loess filters at 0.2 and 0.7 degrees of smoothing, (b) Mean yearlings, Yorlt River, VA, 1950-1982, loess filters at 0.2 and 0.7 degrees of smoothing. exhibit peaks around 1957-1958. 1965 and 1975. and followed a general pattern similar to that of the spat. Rappahannock River The spat pattern in the Rappahannock (Fig. 6a) shows a degree of coherence with the James: high values (two to five) but quite variable before 1955, with a decline through 1961 (less than one), then a significant "recovery" (greater than three) during the mid- 1960s drought, a return to poor set (one to two) by 1970, and finally, a slight increase through 1990. There is also a short re- sponse (1981-1983) to the drought during the eariy 1980s. The yearling abundance patterns parallel that of the spat, exhibiting a decline from 1950 through the early 1970s, followed by a slight recovery (Fig. 6b). Potomac River The Potomac spatfall (Fig. 7) has remained fairly constant since 1950. with short 1- to 3-y responses to the droughts in the 1960s and 1980s. The loess filters show no "1960s decline." only a moderate increase during the 1960s drought, and a subsequent decline through the early 1970s. It is possible that the decline from 1950 through 1972 was interrupted and partially masked during LOWESSO? LOVVESS.0 2 O observations (b) — I \ 1 \ 1 1 1 1 1 — 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 year Mean Yearlings Rappahannock River, VA LOWESS. 0 7 b - 0 0 0 0 LOVVESS.0 2 ot}serva&ons c 4 ^ oo -1 8 3 - \^5^ ~^\° + 0 ^\ ■^ 2 - 060 A 0 " m V") 0 1 - 0 - o I 1 1 1 I 1 1 1 I 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 year Figure 6. (a) Mean spatfall, Rappahannock River, VA, 1946-1992, loess niters at 0.2 and 0.7 degrees of smoothing, (b) Mean yearlings, Rappahannock River, VA, 1946-1977, loess filters at 0.2 and 0.7 de- grees of smoothing. the drought. An apparent recovery is seen from the early 1970s through 1985. Interreef and Intrareef Coherence The coherence of the cumulative abundance of annual spatfall patterns was examined between oyster reefs within river and be- tween rivers by use of the MINITAB Agglomerative cluster anal- yses and Pearson correlation. The analyses were run on James. Rappahannock, and Potomac River reefs. There was an insuffi- cient number of either reefs or unbroken data strings of sufficient length in the York to allow comparisons in that river. The degree to which two time series exhibit the same features of temporal variation can be measured with the simple Pearson correlation coefficient between the two sets of data. Correspond- ingly, a visual comparison can be made from a scatterplot where each year is represented by a point, the (.v. y) coordinates of which are given by the respective observations at the two stations. If the series from the two stations are approximately synchronous, the scatterplot will show the pattern associated with two well- correlated variables. These more complex relationships are conveniently explored Analysis of C. virginica Recruitment 569 Mean Spatfall Potomac River, MD 6 O S _ ° o° 0 _ 4 - o C °°o o o,, 0 0 ° j^"^ ry-"^^ O 8 ^ - __^>_^-Ar— ,^° ° c + o- o° O ' .o 3^^ O r. 2 - O) o 1 o o 0 LOWESSO? 0 - o LO\AESS.0 2 0 observations 00 o 1 ! 1 1 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 year Figure 7. Mean spatfall, Potomac River. MD, 1942-1986, loan filters at 0.2 and 0.7 degrees of smoothing. by use of the cluster analysis technique. The starting point is the calculation of a distance matrix, D. the elements of which, t/y are given by: S| 1B75 - James River o t, 6 " " 8 o n :.,;i 8 • » s / % ■vf' .13 ma "ra ^^^ Correlalion Coefficients James Rjver Stations sl23t s043t s073t sl75t s043t 0.802 s073t 0.767 0.891 sl75t 0.688 0.546 0.692 s016t 0.243 0.367 0.534 0.886 Figure 8. Cross-correlation matrix for spat on James River oyster reefs. where r, is the correlation coefficient between stations / and /. The closest pair of stations, using this distance measure, is combined into a single cluster. The distance matrix is recalculated for the new set of stations (now one less in number), and the closest pair are combined into a cluster. The process can be repeated until a single cluster is produced. The results are best displayed in a dendogram. as shown, for example, in Figure 9. which clearly demonstrates the clustering hierarchy and the presence of two distinct groups of stations with regard to their temporal structure of spatfall counts. James River Figure 8 is a matrix of scatterplots for all 10 possible pairs of the five stations in the James River. The corresponding correlation coefficients are shown in a parallel matrix representation. It is seen that the highest correlation of 0.891 ( / ^ *^ /" Z -'' ^ ^ UJ UJ UJ LJ LJ ' UR ' MR LR R UR Stations mid-nver Stations Figure 11. Agglomerativc cluster analysis of oyster reefs, Rappahan- Figure 13. Agglomerative cluster analysis of oyster reefs. James and nock River. Rappahannock Rivers. Analysis of C. vikgimca Recruitment 571 Potomac River 1 SMS - 1 '"> - 1 1 1 " g" tfo 0 = r^ h h.- ' l„-° ^ " ,. ■ f |.. r-" f 1. ^•» i- L. i... 'i' I i.- ■ tin • ,1» *> ^> iS^ •» « ,,»„1* 9 0' >^ » ,,11^1!. •"■*- — ca. -„ ,^ <™ Correlat on Coefficienis Potomac Rjver Popes Cobb Cedar Heron Swan Jones Ragged Cobb 0.979 Cedar 0.930 0.913 Heron 0-993 0.953 0.821 Swan 0.759 0.747 0.787 0.693 Jones 0.309 0.256 0.3O3 -0.125 0.210 Ragged 0.764 0.550 0.348 0.737 0.337 -0.037 Cornfie Id -0.037 0.090 -0.07B -0.122 •0.125 0.253 0.400 Potomac Stations, James, Rappatiannock Potomac Stations James Rappahanock Correlation Coefficients James 1 Janes 1 BaOP- Jones Cedar Sheepshead Swan Cobb Heron me^nll 0,6?1 meanrapp 0 3i4 0 4se Jones Cedar 0,197 0 436 0,093 0.201 0,2'S 0.303 Sheepshea 0 ?84 0,067 0,266 0.333 0.920 S«an 0.?03 0.148 0.24& 0.210 0.?87 0.776 Cobh 0.261 0.046 0.200 o.2se 0 913 0 964 0,747 Heron 0 330 0.043 0.192 0.12S 0 e?i 0 932 0 693 0 953 Cornfield 0 Oil 0 07? 1? 136 0 ?S3 0 079 0.079 O.IJS 0 090 0.12? Figure 14. Cross-correlation matrix for spat on Potomac River oyster reefs. Figure 16. Cross-correlation matrix for spat on James, Rappahan- nock, and Potomac oyster reefs. Relationship Between Spat and Subsequent Cohort Stages Count.s ofspatfall have been maintained by Virginia and Mary- land since 1946. The original purpose of the spatfall monitonng in Virginia was to provide the state's oyster growers with information on the location and timing of peaks in spatfall to allow them to broadcast shell to receive best the annual "stnke." Over the years, and after the prolonged decline in market oyster landings that coincided with the decline in spat abundance, the annual spatfall report became a forecast for the status of the Virginia oyster har- vest (e.g.. Barber 1991). This relationship was never documented. Spat Versus Yearling The relationship between spat and subsequent cohort stages can be conveniently investigated by obtaining the cross-correlation function (cef) for the spat and the yearling times series. Figure 18a shows the ccf for the James River data. The structure in the func- tion is primarily due to the long-term trend in the data. It appears that, superimposed on this background, there is an enhancement at a lag of 1 y, indicative of the expected relationship between spat density and yearling density in the following year. This relation- ship is revealed more clearly by removing the long-term trend from both data sets and computing the ccf of the residuals (Fig. 18b). The long-term trends are estimated by use of the loess smooth- ing technique. The ccf of the residual series is shown in Figure 18b. It is seen that there is a significant correlation between the series when lagged by 1 y (i.e., spat in year t are compared with yearling in year / -I- 1 ) and that the correlation does not extend Distance Potomac River 091 - 061 - 0 30 - 1 Popes Heron Cobb Ce s ar Sw tations an Rae ged Jo es Com field Distance Potomac Stations, James, Rappahannock ^ 096 - 1 1 064 - 032 - , 1 1 ' ' 1 Stations/Rivers c^" ^"^ ^ Figure 15. Agglomerative cluster analysis of oyster reefs, Potomac Figure 17. Agglomerativecluster analysis of oyster reefs, James, Rap- River, pahannock. and Potomac Rivers. 572 (a) o 9) -0.5 James raw data Cross correlation function Spat — yearling lag (years) Austin et al. (a) 0.5 8 00 York raw data Cross correlation function Spat — yearling Ai^ r Ji -15 -10 lag (years) 10 15 (b) James de-trended data Cross correlation function Spat — yearling 9> lag (years) Figure 18. (a) ccf for James River spat and yearling, (b) ccf for James River spat and yearling (detrended). (b) York de-trended data Cross correlation function Spat — yearling 05 c o S "" m LLiI mn -15 -10 0 j11 10 15 lag (years) Figure 19. (a) ccf for York River spat and yearling, (b) ccf for York River spat and yearling (detrended). beyond 1 y. This can be considered to be a confirmation of the accuracy of designating "yearHngs." Similar results are found in the York and Rappahannock Riv- ers. The ccf values for the raw and detrended data, respectively, are shown in Figure 19 for the York River and Figure 20a and for the Rappahannock River. In both rivers, there is a significant ccf at a lag of 1 y. Having established the presence of a 1-y lag relationship be- tween spat and yearling with no other lags having a significant effect, one may use linear regression between the mean spatfall values for each river, lagged a year, and the mean yearling data. This relationship accounted for a significant percentage of the variation m the James (R" = 0.73; Fig. 2Ia), roughly half in the Rappahannock (R' = 0.48. Fig. 21b). and less than 15% in the York (R- = 0.14, Fig. 21c). The disparity in the coefficient of determination between rivers may be explained in several ways. The James oyster reefs are located in a more geographically compact area, but with a diverse environment strongly influenced by the gravitational circulation (salinity driven deeper "salt intrusion"; Pritchard 1952), which results in a retentive circulation pattern in the lower river (Kuo et al. 1990). It has also been suggested that the proximity of the lower James to the ocean provides a "healthier" environment than the up-bay tributaries (Kuo and Neilsen 1987). Rappahannock spat survival may be less because of summer and late fall hypoxia (Officer et al. 1984, Kuo and Neilsen 1987). Spatfall in the James has a higher chance of retention, survival, and reaching the year- ling stage, whereas survival in the Rappahannock and York is less. Further, repletion efforts (shelling the bottom and seed planting) in these river have been shown to affect results (Ulanowicz et al. 1980. Chai 1988). James River Spat Versus Seed The logical progression for predicting future harvest, with a significant spat-yearling relationship, would be to examine the yearling-seed relation. This was not possible, however, because of the deficiency in the length of the yearling data collection period, which ended in 1984. Further, catch per unit effort (CPUE) data for seed and market oyster are not available until 1983, and then only for the James. This allows only a 1-y overlap, in only one river. Consequently, we examined the James River spat-seed/day Analysis of C. virginica Recruitment 573 (a) Rappahannock raw data Cross correlation function Spat — yearling 0.5 0.0 r Jill -15 -10 10 15 lag (years) (b) Rappahannock de-trended data Cross correlation function Spat — yearling 0.5 - n n H .1. II 1 1 ll.l 1 1 1 1 1 1' 1 1 1 1 1 -15 -10 -5 10 15 lag (years) Figure 20. (a) ccf for Rappahannock River spat and yearling, (b) ccf for Rappahannock River spat and yearhng (detrended). (a) James spat vs yearling (lag 1) (b) Rappahannock spatvs yearling (lag 1) (c) York spatvs yearling (lag 1) Figure 21. (a) Regression of James River spat versus yearling (lag 1 y). (b) Regression of Rappahannock River spat versus yearling (lag 1 y). (c) Regression for York River spat versus yearling (lag I y). relationship. Unfortunately, the short time period of the CPUE data prevented reliable differencing, or detrending, so analyses were conducted with the trend present in the data. Spat and year- ling data were collected by fishery-independent surveys, seed from fishery-dependent commercial harvest data reported to the VMRC by watermen. Pearson correlations were run between log(seed/day) and the spat value lagged 1 through 4 y as a mean of narrowing the field of observations for subsequent regression analyses. Significant correlations were found at lags of 2 and 3 y (Table 2), and to a TABLE I. Station data flies (stations selected for analysis contain 50% of all available information). Station Size of File (bytes) Cumulative (bytes) River Station Name S073 SI 75 S180 S181 SOOl S067 S190 S043 SI79 S050 SI09 S016 S123 SOU 2.772 2.772 2,772 2,772 2,583 2,583 2.457 2.268 2,268 2,142 2,142 2,016 2,016 1,953 2,772 5,544 8.316 11,088 13,671 16,254 18,711 20,979 23,247 25,389 27,531 29,547 31.563 33.516 James James Rappahannock Rappahannock York Rappahannock Rappahannock James York Piankatank York James James Rappahannock Horse Head Wreck Offshore Moratlico Bar Drumming Ground Aberdeen Rock Hog House Bar Smokey Point Deep Deepwater Shoals Bell Rock Ginney Point Pages Rock Brown Shoals Point of Shoals Bowlers Rock 574 Austin et al. TABLE 2. Pearson correlation coefficients and linear regression of James River spat versus seed/day. Pearson Correlations Linear F P :egression Parameter Log seed/day Spat Spatl Spatl Spat3 R"(%) Spat Spatl Spat2 Spat3 Spat4 -0.130 0.097 0.676 0.681 0.517 0.098 0.035 0.107 0.282 0.125 0.061 0.104 0.097 0.066 0.103 Equations Log seed/day = 2.610 + 0.078 spatl Log seed/day = 0.606 + 0.507 spat2 Log seed/day = 0.565 + 0.520 spat3 Log seed/day = 1.040 + 0.408 spat4 0.790 0.032 0.30 0.126 0.9 45.8 46.4 26.7 lesser degree at 4 y. Regressions of seed/day and spat lagged 1 to 4 y were also run. The 2- and 3-y lag was found to be significant at the p = 0.05 level (Table 2; Fig. 22). 4 5 spat (lag 2 years) 4 5 spat (lag 3 years) Figure 22. Regression for James River spat (lagged 2 and 3 y) versus James River market oysters. Spat Versus Market Oysters The relationship between James River spat and subsequent years' market oyster landings was examined. As with the spat/seed analysis, only James River spat/market was examined because the James River landings are not "contaminated" by oyster repletion and because the spat and market/day are from the same river. The short CPUE data series (market/day) precluded differencing. Pear- son correlations were run for spat against market/day 1^ y later. None were significant (Table 3), although there was a negative correlation between market/day and spat 2 y earlier, which is probably an artifact of the short, nondetrended market data. Krantz and Merritt ( 1977) found their best correlation between spat and commercial harvest at a lag of 6-8 y and cited this as further evidence to ". . . sustain the theory that a period of suc- cessive years of low spat set will require between six to eight years before the period of poor recruitment is reflected in the commer- cial harvest." Ulanowicz et al. (1982). using a multivariate anal- yses, found a correlation between spat and seed at a lag of 4 y, and using cross-correlation analyses, found a peak in the correlation between spat and commercial harvest at a 9-y lag. This period, they speculated, could be due to a ". . . possible natural oyster cycle . . ." or an ". . . unexplained environmental variable." These 4-, 6-. 8-. and 9-y lags found by Krantz and Merritt and Ulanowicz et al. may be artifacts of the cross-correlation because "... interpretation of the sample cross-correlation function can be fraught with danger unless one uses the prefiltering proce- dure . . ." (Chatfield 1989). Neither study detrended the raw TABLE i. Pearson correlation coefficients for James River market oysters/day versus spat lagged 1^ y. Lag Year r Spall Spat2 Spat3 Spat4 -0.283 -0.696 -0.490 -0.125 Analysis of C. virginica Recruitment 575 TABLE 4. Pearson correlation coefTicients and linear regression coelTicients and equations fur James River seed/day versus market/day. Pearson Correlations Parameter Log market/day Seed I /day Seed2/day Seed3/day Seed4/day Linear Regression P R' (%) Seed I /day -0.063 Seed2/day 0.160 0.732 Seed3/day 0.513 0.387 Seed4/day 0.716 0.028 Equations Log market/day = 1.53 + 0.238 log seed3/day Log market/day = 1.17 + 0.355 log seed4/day 0.802 0.554 0.820 0,194 26.3% 0.071 51.2% data, so it is quite possible that the 4- to 9-y lags that they found are artifacts of this lack of differencing. Their multivariate analysis revealed that spat densities and seed planting accounted for 56% of the variation in commercial harvest. The removal of significant volumes of seed from the James River, and their transplantation to the Rappahannock and Potomac Riv- ers, has no doubt affected the statistical results of our spat versus seed and market analysis. James River Seed Versus Market Oyster The James River abundance of seed was analyzed relative to James River market oyster CPUE. Normally, one would expect that the market oyster catch is composed of several year classes or cohorts. This is true for the James (Mann, unpublished data); however, with the current level of fishing pressure, depleted Market/day vs seed/day (log transform) lags 0, I, 2, 3, 4 years ■ J ° <- ' i , s o ' ' ° ° °o 1" • °° ° % " ICBOeaVAr) to9*matag 1 S" J, ° ■ ■ " Stocks, small minimum size limit (3". 76.2 mm), and slow variable growth rates, it is likely that the commercial harvest, although composed of several year classes, is supported primarily by only a year class or two, most probably, age four. Nevertheless. Pearson correlations were run on log( market/day) by log( seed/day) lagged 1-4 y (Table 4). Because of the short overlap period with effort (boat days, 1983-1994), the data were not differenced. There was a significant correlation between market oyster and seed, lagged by 3 and 4 y (0.513, 0.716), and a slightly significant regression (p = 0.071, R- = 51.2%) with seek laged 4 y (Table 3; Fig 23). This relationship appears to be fortuitous and is probably due to the strong downward trend in seed after 1985 and the pulse of market landings in the later 1980s, also followed by a dramatic decline. Predictions With Spat Because spat (age "zero-plus") show a statistical relationship to seed (age two and three), intuitively, one might expect that there would be a relation between seed (2-3 y) and market oyster (age three and four) a year later. However, there is no significant re- lation between spat and any market size, and the seed-to-market relationship is between the age two and three seed and age six to seven market oyster. It is our conclusion that spat abundance can be used to predict the abundance of subsequent yearling oyster abundance and can form the basis for a method of predicting abundance of seed Market and Seed Oyster James River, VA Figure 23. Regression for James River market/day versus lagged seed/ day. 1963 1969 1975 1981 1987 1993 - seed ^ market I Figure 24. Total market and seed oyster harvest, James River. VA. 576 Austin et al. Total Harvest Boat Days James River, VA 25000 ^20000 to 2 15000 + (0 o ^ 10000 TO ° 5000 1960 1965 1970 1975 1980 1985 1990 1995 Year Figure 25. Number of boat days, James River, VA. (CPUE) 2-3 y later. It does not appear, at this time, that spat can be used to predict future market oyster harvest. It may be that when the catch/day data set is longer, it will be possible to make a correlation after detrendmg. Further, although there is an appar- ent relation between seed CPUE and the market CPUE 4 y later, we feel that this may be due more to the overall trend of the data rather than to biologic cause and effect. Further, the multiple cohorts in the market catch and problems with CPUE data for seed and market oyster make this examination questionable. We will discuss the additional problems with seed and market data as to how they relate to CPUE when calculated with boat days. With this in mind, any examination of seed or market landings must be made with caution. James River Seed and Market Harvest The VMRC has maintamed monthly harvest statistics since 1963 for seed, and market (currently, 3", 76.2 mm) oyster, and since 1983 the number of boat days fished in the James. Figure 24 (Table 4) depicts the annual harvest of seed and market oyster in the James River since 1963. Figure 25 shows a dramatic increase Market Oyster Harvest James River, VA 1960 1965 1970 1975 1980 1985 1990 1995 Year -^ market • cpue Seed Oyster Harvest James River, VA Figure 27. Seed oyster harvest and CPUE, James River, VA. in boat days (effort) m the James through, and peaking in. 1988 and an equally dramatic decline thereafter. This variation is due to the scarcity of oyster bay-wide, except in the James, a subsequent migration of the watermen from the less productive waters of the TABLE 5. James River seed and market oyster harvest 1963-1995 (all data from VMRC). Seed Market Boat Year (x 1,000 Bu) (xlOOBu) Seed/Day Market/Day Days Figure 26. Total market harvest and CPUE harvest, James River, VA. 1963 844 175.7 1964 830 417,4 1965 424 450.0 1966 611 487.9 1967 533 167.0 1968 484 182.0 1969 487 157.7 1970 264 143.8 1971 459 170.8 1972 381 129.7 1973 396 27.4 1974 373 186.3 1975 317 61.6 1976 441 14.6 1977 420 3.3 1978 350 13.2 1979 420 42.7 1980 350 68.4 1981 214 136.0 1982 406 21.5 1983 445 16.1 62.8 0.2 7,087 1984 346 48.8 45.9 0.6 7,533 1985 410 21.5 54.4 0.3 7,537 1986 277 28.8 49.2 0.5 5,625 1987 199 341.4 12.6 2 "^ 15,754 1988 136 297.2 6.4 1.4 21,305 1989 135 146.2 9.6 1.0 14,027 1990 51 68.2 5.2 0.7 9,810 1991 55 36.5 8.2 0.5 6,698 1992 54 25.6 13.4 0.6 4,032 1993 95 20.0 35.2 0.7 2,698 1994 75 5.5 43.7 0.3 1,715 1995 126 17.7 36.0 0.5 3,500 Analysis of C. vircinica Recruitment 577 Mean James spat 9 - 8 - 7 - • 6 - .1 „ 5 - CO '-r-^^^rr^^'^' 3 - ■ 2 - . , 0 - 17 18 19 20 21 1 Spring mean water temperature Rappahannock - upriver 4 — • 3 - ; Q. 2 - =) C S ' --^--^--CL^ E * '~ — — ^_^ Y = 8 46964 ■ 1 4S657X *- 0 - . . > • •• • " * R-Squared = 0 102 -2 - 45 50 55 logsumm Figure 28. Regression of mean James River spat versus spring VIMS pier temperatures. Figure 29. Regression of upper Rappahannock River spatfall versus summer river flow. Rappahannock and Potomac Rivers mto the James for both seed and market oysters, and a dechne m effort as catch dropped off. Figure 26 depicts the market oyster and CPUE since 1983. with data derived from the market oyster harvest and both days (Bush- els Market oyster/boat days). Reduced oyster stocks and active management by the marine Resources Commission combined to result in a post- 1990 reduction in effort. Both harvest of market oyster and CPUE of market oyster parallel boat days. This is because watermen would rather focus their efforts toward harvesting $30/Bu of market oyster, than $4/ Bu of seed. After 1990, however, as stocks of market oyster became seriously depleted, significant effort was redirected to- ward the harvest of seed, the remaining resource. A quota system for seed was introduced in 1993-1994, permitting the harvest of 80 kBu. but the limit was increased at the watermen's insistence to 120 kBu in 1994-1995. Although CPUE for seed increased in 1993-1995, the total seed harvest has remained relatively stable since 1990 (Fig. 27). A significant note of caution should be introduced. Although the number of boat days has been recorded monthly since the 1982-1983 season, they were not separated between seed harvest days, market harvest days, and which days were a split between the two activities. In other words, of the 5,625 boat days in 1986. it is not possible to determine how many of these were spent harvesting seed and how many were spent harvesting market oys- ter. Consequently, the CPUE calculations for seed and market were made with the unlikely assumption that equal numbers of days were spent on each fishery. In short, although the calcula- tions have been made, we would not place great reliability on them TABLE 6. Regression analysis for James and Rappahannock upriver spat abundance versus spring and summer river flow. River Season R- James River Spring 0.089 0.034 James River Summer 0.018 0.074 Rappahannock River Spring 0.102 0.052 Rappahannock River Summer 0.023 0.102 because the data are so ""spongy." This results in the Fisheries Management Axiom: Are '"spongy"" better than none? Other factors may influence the results here in a way that cannot be estimated. The first is that the James River is the source of seed for the Virginia repletion program that transports seed oyster from the James to nonproducing areas of the Virginia trib- utaries. This movement of seed may result in changes in abun- dance both in the James (Downward) and the other rivers (upward) that are not reflected in our count data. The second factor is the spread of disease, which has been responsible for much of the mid- and late- 1980s decline in market oyster (Bureson and Ragone Calvo 1996). After reaching 19—45 mm, when 2-3 y of age, the seed oyster in the lov^er, more saline regions of the James River become susceptible to the diseases MSX (Haplospohdium nelsoni) and Dermo (Perkinsiis murimis). Burreson and Ragone Calvo (1996) have found this to be particularly severe on Wreck Shoal in the James River, where mortality has been lOO^f for several years. The removal of seed and market oyster from the stock by either disease or repletion will obviously affect our results, but we are unable to estimate to what degree this has occurred. It is our conclusion that the seed and market CPUE data, as currently col- lected, cannot be used to examine the effect of seed abundance on subsequent market landings. Summer Palmer Drought Index Tidewater Virginia 1960 1970 1980 Year Figure 30. Summer PDI, Tidewater, VA. 578 Austin et al. Smoothed upnver spat 1 - f f!app- ?\ 0 - '%u \ 4- + -1 - James ij J 1950 1960 1970 1980 1990 Year James Upriver 1.0 - 0.5 - 0.0 - -0.5 - -10 - -15 - smoothed spnng j-js >-•■*" '^ ^ smoothed spat 1950 1960 1970 1980 1990 Year Figure 31. Detrended, smoothed upriver spatfall for James and Rap- pahannoclt (Rapp.) Rivers. Figure 33. Detrended, smoothed upriver James spatfall and de- trended, smoothed spring PDI. Relation of Spat to Its Physical Environment Most marine organisms, particularly those attached to the bot- tom, are susceptible to fluctuations in the physical environment. Numerous articles addressing these oyster-environment relation- ships have been published (Ulanowicz et al. 1980. Haven 1982, James River 0 5 - n n 1 ilh .ll 1 1 '1 !'■ -0.5 - -10 -- 1 1 1 1 10 15 lag (years) 20 25 Chai 1988. Austin et al. 1993). Generally, they have pointed to temperature and salinity (or its proxy, river discharge) as the con- trolling physical variables. Temperature Effects on Spat The water temperature data measured at the VIMS pier at the mouth of the York River constitutes an almost continuous data set since 1952 and was used as surrogate data for all of the rivers. The effects were examined of mean spring temperature (May through July) and mean summer temperature (July through September) on the mean spat from the James and from the Rappahannock Rivers. In no case did the value of R^ exceed 2.1%, and none of the regressions were significant (Table 5). As an illustration of the lack of relationship, the data and regression line for the "most significant" (p = 0. 18) regression between James River spat and spring temperature are shown in Figure 28. The conclusion is that the water temperature during the spring and summer preceding the spat measurement has minimal effect on the spatfall. River Discharge Effects on Spat River discharge is monitored by the U.S. Geological Survey and the NOAA Office of Hydrology. We used data from the mon- itoring stations located on the fall line of the Rappahannock and 1 n Rappahannock River James upnver 5 0,5 - g ^ on ' L ll h. 1.0 - 0.5 - 0.0 - -0.5 - -1.0 -1.5 - ^^smooth summer ^ /^ \* smcWh spat V f 8 -0.5 - 1 1 'II c ) 1 5 1 1 1 10 15 20 lag (years) 2 1950 1960 1970 1980 1990 Year Figure 32. cff for detrended and smoothed James River and Rappa- Figure 34. Detrended, smoothed upriver James spatfall and de- hannuck River spatfall. trended, smoothed summer PDI. Analysis of C. virginica Recruitment 579 James Upriver 1 - Palmer summer rate 9\ 0 - 7 u -1 - A f spat 1950 1960 1970 1980 1990 Year Rappahannock upnver 1.0 - c^sfnooth spat 0.5 - 0.0 - \lJ\\^T\ -o.s - ^^"^ 4- *smooth summer -1.0 - -1.5 - 1950 1960 1970 1980 1990 Year Figure 35. Detrcnded, smoothed upriver James spatfall and de- trended, smoothed summer period of maximum rate of change in PDI. Figure 37. Detrended, smoothed upriver Rappahannock spatfall and detrended, smoothed summer PDI. James Rivers and selected "upriver" stations by using agglomer- ative cluster analysis characterizations of the oyster reefs. It was expected that stations furthest up stream would be those most likely to reflect fluctuations in stream flow (Haven 1982). These included the Group 1 James River reefs (Decpwater Shoals, Horse head, and Point of Shoals) and both Morattico and Bowlers Reefs in the Rappahannock River. We looked at both spring (May to July) and summer (June to September) mean discharge patterns for the James and Rappahan- nock Rivers and regressed them (log flow) against the mean spat- fall abundance for the two upriver populations. It is obvious from the results in Table 6 that with the exception of the Rappahannock summer flow (Fig. 29), spring/summer river discharges alone did not produce a significant variation in spatfall patterns. Andrews et al. ( 1939) noted that the significant 1957 spat set was largely wiped out during the 1958 winter-spring freshets. Although the fall survey count showed a large set in 1957, mor- tality was high during the following May to June period, when the previously overwintering dormant spal became active in the low- salinity James. They also reported that although this occurred in the James, they did not notice a similar effect in the Rappahan- nock. Haven (1982) reported that the prolonged periods of low salinity during the fall, winter, and spring of 1979-1980 produced extensive mortalities of the 1979 set in the James River. Yearling were affected to a lesser degree, and market size oyster exhibited the lowest mortality. This is reflected by a lower fall count of yearling oyster in the James River in 1980 (Fig. 4b). PDI Relation to Spat Abundance Neither temperature nor river discharge (proxy, salinity) data gave significant relationships with spat abundance, in spite of historic reports and an intuitive assumption that they should. In search of an alternative environmental variable, we considered the PDI, a combination of air temperature, precipitation, and soil type as a possible integrated environmental signature. The index is in standard usage by climatologists and is published monthly by the Office of the Virginia State Climatologist at the University of Virginia. Precipitation data alone do not always reflect river dis- charge and. consequently, salinity, because it is often the rate of the precipitation that influences the amount of runoff that ulti- mately results in river discharge. Rain soaks in, while rain showers often exceed the soil's absorption capacity and result in runoff to the creeks and rivers. The PDI is computed for four areas of the state, depending on the temperature, precipitation, and soil type regimens. We considered that the Tidewater index was appropriate for this study. 0.6 0.4 § 02 8 00 -0.2 H -0.4 -15 -10 -5 10 15 20 10 15 -5 0 5 lag (years) lag (years) Figure 36. Cross-correlation of James River spat and smoothed sum- Figure 38. Cross-correlation of spring and summer Rappahannock mer period of maximum rate of change in PDI. River spat and PDI. 580 Austin et al. Rappahannock upriver 12 - • « • ^^^'-''''^ spat • • • • Y=033.052X R-Sqiflrad = 0 78 -08 - -15 ■10 -05 00 05 10 Palmer index - 5 year lag Rappahannock upriver ti - • ^ • ^„.'^^ ^^^» \ 02- • • • • V=035*057X -08 - -15 -10 -05 00 05 10 Palmer Index - 6 year lag Figure 39. Regression of Rappahannocli River spat and PDI lagged 5 and 6 y. The PDI is a negative or positive deviation from normal. A positive index represents wet conditions (e.g., 1979. >4.0), and a negative index represents dry or drought conditions (e.g., 1986. > — 3.5). Figure 30 depicts the summer index, which is reason- ably representative of both the spring and the summer. The "dry" 1.0 05 -0.5 -1.0 .1 1 1 1 1 ' 'I 1 -11 -15 -10 0 10 15 lag (years) Figure 41. Cross-correlation between Rappahannocit River spat and summer period of maximum Rate of change in PDI. or drought conditions of the mid- 1 960s and mid- 1 980s are readily apparent, as is the trend away from drought to "wet" conditions during 1966 through 1979. There were no "wet" periods of over 2 y during the period of measurement, 1950 and 1994. We computed a spring (May to July) and a summer (June to September) mean index. The ccf between the summer and spring PDI was 0.845, and the regression coefficient was 0.71. As discussed above, if there is a relationship between fall spat- on-shell and river discharge (i.e., salinity) and/or temperature, it should be most readily apparent at the stations furthest upriver. Using the agglomerative cluster analysis characterizations of the oyster reefs, we picked the James and Rappahannock "upriver" groups for analysis. Strong long-term trends in spat data, partic- ularly those following the post- 1960 decline, were apparent (Figs. 4-7). Cross-correlation was the analysis tool planned for exploring the relationship between spat and river discharge. As we have observed earlier, the results of this type of analysis can be severely corrupted by the presence of long-term trends (Chatfield 1989). Such trends were therefore removed by use of the loess filter in MINITAB. with the adjustable parameter set to remove all but the lowest frequencies (parameter value = 0.7). The residuals from this smoothing constitute the detrended data. The random year-to- Rappahannock Upriver 1.0 - /\ ®P^' 0.5 - \ 0.0 - A \i ''\j"*A -0.5 - Vv ^ ' \ 1950 1960 1970 1980 1990 Year Rappahanock upriver 1 2 - 02 - t (0 .08 - "--^^^.. Y = 0 124813-1 44892X R-Squared = 0 82S -05-04-03-02-01 00 01 0.2 0.3 04 OS Palmer rate -- 1 year lag Figure 40. Detrended, smoothed upriver Rappahannocic spatfall and Figure 42. Regression of detrended, smootlied upriver Rappahannock detrended, smoothed summer period of maximum rate of change in spatfall versus detrended, smoothed summer period of maximum rate PDI. of change in PDI, lagged 1 y. Analysis of C. virginica Recruitment 581 Rappahanock upriver 1 - to- (0 -1 — V>v.- ' Y = 0 107691 ■ 1 47726X R-Squared = 0 B33 -0 5 -0 4 -03 -0 2 -0 1 00 01 02 03 04 0 Palmer rate -- lag 2 years Figure 43. Regression of detrended, smoothed upriver Rappaliannock spatfall versus detrended, smoothed summer period of maximum rate of change in PDI lagged 2 y. year fluctuations in these data were smoothed by a further appli- cation of the loess filter with a parameter of 0.3. The overall effect of these procedures was a bandpass filtering of the original data whereby long-term trends and short-term fluctuations were re- moved. Figure 31 shows these smoothed data for both rivers through time. By visual inspection, the James River spat exhibited an 1 1- to 12-y cycle and the Rappahannock exhibited a 15-y cycle. This observation is confirmed by an inspection of the ccf ( Fig . 32 ) . James River Figure 33 shows the James upriver smoothed spat and spring PDI, and Figure 34 shows the spat and summer PDI. In each case, it is apparent from inspection that the summer precipitation deficits are more profound than the spring, but that both seasons move in synchrony. Also, visual inspection shows that spat and PDI fluc- tuate out of phase. If, however, one considers the period of great- est change in PDI (APDI). as opposed to the actual values, it is apparent that they are in phase (Fig. 35). The peaks and valleys of the spat data are in phase with the periods of greatest APDI. Regression of the summer APDI against spat yielded an R" of only 14.2%. However, when the period of greatest APDI was cross- correlated with spat, the greatest correlation was found at a lag of 4 y (-0.372) (Fig. 36). Rappahannock River The Rappahannock smooth spat and summer PDI demonstrated a greater degree of visual synchrony (Fig. 37) than the James, and the R^ was 26.9%. Cross-correlation analyses showed a significant lag of 6 y (0.880) between spat and summer PDI. and a 5-y lag (0.817) with the spring PDI (Fig. 38). The R" for the regression between spat and the 5- and 6-y lag of the summer PDI were 0.77 and 0.90 (Fig. 39). When the rate of change in the PDI (APDI) (Fig. 40) was cross-correlated with spat (Fig. 41). the greatest correlation was found at a lag of I and 2 y ( -0.881 and - 8.62). R* for the lags were 0.825 and 0.833. respectively (Figs. 42 and 43). The re- sponses of the spatfall to the changes in the PDI are reflected both in the 1960s, as conditions evolved from "damp" to "drought." and in the more prolonged "drying" period of the mid-1970s to mid-1980s, as the spatfall reflect a short and a longer period of increased set (Fig. 40). The James River, because of its proximity to the mouth of the Chesapeake Bay. is more under the influence of oceanic-salinity gravitational circulation (Pritchard 1952. Neilson and Kuo 1989) than the Rappahannock. This circulation regimen has been sug- gested in the past (Austin et al. 1993) to be the cause of some interriver variations in oyster condition index. As such, it is not unexpected that the upper Rappahannock showed the greatest re- sponse to fluctuations in freshwater input. It must be pointed out. however, that Andrews et al. (1959) found no such James versus Rappahannock differences when examining extremely low flow patterns. Spat and PDI Linkages Statistically, high cross-correlations and/or regression coeffi- cients between PDI and spatfall at 7 or 8 y do not make ready biologic sense. Yet, it is coincidence that this is the same lag period found to be statistically significant by Krantz and Merritt (1977) and Ulanowicz et al. (1980) for spat to harvest? Stepwise multiple regression by Ulanowicz et al. also showed that "drought episodes." cumulative (sustained) excessive salinity, extreme rainfall during the previous season, and harvest all caused direct variations in spat density. In this study, however, the "depth of the drought" or "peak period of rainfall/runoff" might not be expected to show a direct cause and effect with spatfall because the long lag is unexplained biologically. On the other hand, if one considers the period of greatest PDI change (APDI), that period when the environment passes from one temperature/precipitation regimen to another, it makes biologic sense that the populations, after a lag. will begin to show change; then, change will occur rapidly as the population shifts toward equilibrium with the "new" environment. The cyclic nature of the physical (PDI) re- sults in rapid and cyclic changes in the spatfall. Only during the extended drought of the early- to mid-1980s did the spatfall rates have a chance to equilibrate. Allen et al. (1977) and Legendre et al. (1985), looking at succession of species within a community, said that ecological succession evolves in steps, instead of smoothly, shifting from one structure to another, produced by intermittent shifts in the environmental structure. CONCLUSIONS Spatfall in the Virginia tributaries to the Chesapeake Bay and the Potomac River show a pattern of declining set from the late 1949"s through the early 1970s, followed by a moderate recovery after 1975. The decline is not apparent in the Potomac. Counts of yearling oyster follow a similar pattern, except in the York River, where they increased, followed by a steady decline. Patterns of spatfall tended to partition into upriver and downriver clusters. Spatfall levels, as indexed by counts-on-shell, can be used to predict the abundance of seed oyster 2-3 y later, but are not a good predictor of market oyster abundance. This lack of a predictive spat-market capability may be, in part, the result of the movement of seed through the oyster repletion program. Although spat did not show a direct statistical relation to tem- perature or salinity, there was a significant relationship with the PDI, particularly when the index was shifting from wet to dry or vice versa. We attribute this to a shift in the environmental struc- ture of the rivers and the response of the oyster recruitment. 582 Austin et al. Allen, T. F., S. M. Bartell & J. F. Koonce. 1977. Multiple stable con- figurations in ordination of phytoplankton community change rates. £ra/og.v 58:1076-1084. AndrewsJ. D..D. S. Haven & D. B. Quale. 1959. Freshwater winterkill of oysters in the James River, VA. 1958. Proc. Natl. Shellfish Assoc. 49:29-49. Austin, H. M., D. S. Haven & M. S. Mustafa. 1993. The relationship between trends in a condition index of the Amencan oyster. Crassos- trea virginica. and environmental parameters in three Virginia estuar- ies. Estuaries 16:362-374. Barber, B. 1991. Oyster Spatfall in Virginia Waters, 1990 Annual Sum- mary, VIMS Sea Grant Marine Resource Special Report, Feb. 1991, 13 pp. Burreson, E. & L. Ragone Calvo. 1996. Epizootiology of Perkinsiis man- mts disease of oysters in Chesapeake Bay. with emphasis on data since 1985. y. Shellfish Res. 15:17-34. CBSAC. 1988. Stock Assessment Plan. Chesapeake Bay Stock Assess- ment Committee. Chesapeake Executive Council. Chesapeake Bay Program, Agreement Commitment Report July 1988. 66 pp. CEC. 1994. Chesapeake Bay Oyster Management Plan Chesapeake Ex- ecutive Council, Chesapeake Bay Program. Agreement Commitment Report. Annapolis, MD. Chai. A. L. 1988. Vanability of Amencan oyster recruitment and harvest and blue crab distnbution in the Maryland portion of the Chesapeake Bay. M.S. Thesis. Chesapeake Biological Laboratory. University of Maryland. 248 pp. Chalfield. C. 1989. The Analysis of Time Senes. An Introduction. 4th ed. Chapman Hall, London. 241 pp. Cleveland, W. S. 1979. Robust locally weighted regression and smoothing scatterplots. J. Am. Stat. Assoc. 74:829-836. Cleveland. W. S. 1993. Visualizing Data. Hobart Press, Summit, NJ. 360 pp. Hargis, W. J. & D. S. Haven. 1988. Rehabilitation of the troubled oyster industry of the lower Chesapeake Bay. J Shellfish Res. 7:271-279. LITERATURE CITED Haven. D. S 1982. The impact of low salinities on James River oyster population 1979-1980. VIMS. Special Scientific Report in Applied Manne Science and Ocean Engineenng (SRAMSOE). No 258. 18 pp. Haven. D. S. & L. W. Fritz. 1985. Setting of the American oyster Cras- sostrea virginica in the James River. Virginia. USA: temporal and spatial distnbution. Mar. Biol. 86:271-282. Haven, D. S.. W S. Hargis, Jr. & P. C. Kendall. 1978. The oyster in- dustry of Virginia, its status, problems and promise. 2 Ed. Va Inst Mar Sci Spec Papers in Manne Sci, No 4:1-1024. Hidu, H. 1969. The Feasibility of Oyster Hatchenes in the Delaware- Chesapeake Region. Proc. On Contrib. On Artificial Propagation Of Commercially Valuable Shellfish. College of Marine Studies. Univer- sity of Delaware. Newark. NRI Cont. No. 396. Krantz. G. E. & D. W. Meritt. 1977. An analysis of trends in oyster spat set in the Maryland portion of the Chesapeake Bay. Proc. Natl. Shell- fish Assoc. 67:53-59. Kuo. A. & B. Neilson. 1987. Hypoxia and salinity in Virginia estuaries. Estuaries. 10:277-283. Newell, R. & B. Barber. 1990. Summary and recommendations of the oyster recruitment and standing stock monitoring workshop, Horn Point Lab, 6-7 Nov. 1990. Maryland Department of Natural Re- sources, 32 pp. Officer, C. B.. R. B. Biggs. J. L. Taff. L. E. Cronin & M. A. Tyler. 1984. Chesapeake Bay anoxia: origin, development, and significance. Science 223:22-27. Pntchard. D W. 1952. Salinity distribution and circulation in the Chesa- peake Bay estuarine system. J. Mar. Res. 11:106-123. Richkus. W. A.. S, J, Nelson & H. M. Austin. 1992. Fisheries assess- ment and management synthesis: lessons for Chesapeake Bay. Chapter 4. mCRC. pp. 75-1 14. Ulanowicz, R. E.. W. C. Caplins & E. A. Dunnington. 1980. The fore- casting of oyster harvest in central Chesapeake Bay. Estuar. Coast. Mar. Sci. 11:101-106. Journal of Shellfish Research. Vol. 15, No. 3, 583-587, 1W6. EPIZOOTIOLOGY OF THE PARASITE, PERKINSUS MARIN US (DERMO) IN INTERTIDAL OYSTER POPULATIONS FROM LONG ISLAND SOUND DIANE J. BROUSSEAU Biology Department Fairfield University Fairfield. CT 06430 ABSTRACT The receni reported occurrence of Perkinsiis maniuis in oysters from Long Island Sound prompted this study of the epizootiology of the parasite in this region. The monthly prevalence and infection intensity of P. marinus were determined for three intertidal populations of eastern oysters, Crassostrea virginica, during the period of 1993-1996. Total numbers of infected oysters were highest at the Bndgeport site, followed by the Westport and Milford locations. Disease prevalence was greater in 1995 than in 1994 at all sites studied. Numbers of infected oysters and parasite burdens peaked in the late fall and declined dramatically in the winter/eariy spring. Prevalence was highest (20-100%) at the Bridgeport, CT, site in every month sampled, and the weighted prevalence (Mackin's scale: t)-5) reached a maximum of 3.2 in November 1994. Temperature and salinity data available fiir the Bridgeport location indicate that conditions were reportedly favorable for parasite proliferation (S!20°C: >10 ppl) from June through September. KEY WORDS: Perkinsus marinus. Dermo, prevalence, intensity. Long Island Sound, temperature INTRODUCTION PerkiiLsii.s marinus (commonly known as "Dermo"), a proto- zoan pathogen of the Apicomplexan group, is generally considered the most serious oyster pathogen in Chesapeake Bay and the Gulf of Mexico. P. marinus was first identified as a disease -causing agent in the eastern oyster in the Gulf of Mexico in 1947 by Maekin et al. ( 1950), and its occurrence was quickly documented throughout the southeastern United States (Ray 1954). By 1949, this pathogen was discovered in the lower Chesapeake Bay (An- drews and Hewatt 1957, Andrews 1988) and in the mid- to late 1980s, it began to spread to locations farther and farther north in the estuary. P. marinus is now present on nearly all oyster bars in Virginia and Maryland and has recently become established in Delaware Bay (Smith and Jordan 1993. Ragone Calvo and Bur- reson 1995). This apparent range extension northward has contin- ued, and between 1991 and 1992. infected oysters were found from Long Island Sound to Cape Cod. MA. By 1995. the infection of oysters by species of the genus Perkinsus (probably P. marinus) had been reported as far north as the Damariscotta River in Maine (Kleinschuster and Parent 1995). Regional differences in the seasonal cycle of P. marinus have been well documented. They have been attributed largely to vari- ations in seasonal temperature patterns between northern and southern limits of the parasite's range. In the south, the parasite exhibits high prevalences throughout the year, declining very little during the winter months (Andrews and Ray 1988; Soniat and Gauthier 1989, Crosby and Roberts 1990. OBeim et al. 1994). In more northerly waters, epizootics display a more seasonal pattern in which the prevalence and intensity of the disease increase in the late spring, peak in the late summer/fall, and decline over the winter/early spring months of the year (Andrews and Hewatt 1957. Andrews and Ray, 1988. Burrell et al. 1984). A detailed under- standing of the periodicity of Dermo disease in areas in which it has been recently reported, however, is not yet available. Since the first reported cases of P. marinus in Long Island Sound appeared in the early 1990s, the disease has reached epi- zootic levels in numerous locations throughout Long Island Sound, affecting both cultured and wild stock (Brousseau et al 1994. Brousseau 1995. Brousseau unpubl). To date, information about Dermo disease in this area has been generated largely by work on intensively cultured commercial beds. Oyster culling and harvesting activities, however, subject these "populations" to constant disturbance, making it difficult to study the natural course of disease in this setting. This investigation was initiated to deter- mine the epizootiology of P. marinus in western Long Island Sound in three undisturbed natural populations. The data presented here represent the first long-term study (31 mo) of diseased oysters in this region. MATERIALS AND METHODS Oysters {Crassostrea virginica) were collected from three in- tertidal sites in western Long Island Sound: Milford Pt.. Milford, CT; Black Rock Harbor, Bridgeport. CT; and Saugatuck River, Westport. CT (Fig. 1). Oysters from the Milford site were col- lected monthly from September 1993 to August 1994. The CONNECTICUT / (K) miLFORD POINT (b) black bock HAKBOR (c) saugatuck, wistfobt I- MIDDLE GROUND LONG ISLAND SOUND NORTHPORT' NEW YORK Figure I. Map of Long Island Sound showing the geographic locations of the three collection sites: (Al Milford Pt., Milford, CT: (Bl Black Rock Harbor, Bridgeport, CT; and (C) Saugatuck River, Westport, CT. 583 Mllford, CT • u c » > 9 111 LU o z < > a. Q. Q UJ h- X o UJ 5 ^zo"u-5<5->;; « LU O z LU _) < > LU QC a. a LU I- X g LU 5 Wastpon. CT WsslpoTi, CT o u c « n > LU o z LU -J < > UJ a. a. o LU I- X g LU 5 i § g r, S g 2 S :: g ~ S R ' ^ r\j230 (j-m for C. virginica and >280 p.m for C. gigas) and fully developed eyespots on both sides of larvae were also used as indications of larval competence (Pitt and Coon 1992). Metal Toxicity Tests Toxicity tests were designed to demonstrate how three stages (swimming, searching, and metamorphosis) of larval set were af- fected by elevated concentrations of metals. Experiments were conducted in plastic tissue culture plates (24-well; Falcon No. 3047) with filtered (0.2 |xm pore size) Marine Biological Labora- tory (MBL) artificial seawater. which consisted of 423.0 mM NaCI. 9.00 mM KCl. 9.27 mM CaCK, 22.94 mM MgCU, 25.50 mM MgSOj, and 2.15 mM NaHCO,. The MBL seawater has a salinity of 37 g • kg ~ ' and was used in all tests of C. gigas larvae. Two-thirds-strength MBL seawater has a salinity of 24.7 g ■ kg" ' and was used in all tests of C. virginica larvae. These media were adjusted to pH 8.3 with 5N NaOH and were supplemented with 100 p,g • niL" ' of antibiotics as mentioned above. Copper or zinc were introduced as concentrated stock solutions (chloride salts) into the test media in which the larvae were exposed. Total metal concentrations tested were 0. 100, 500 and 1 .500 ppb (|j.g/L). and all treatments were run in triplicate. Metal ion speciation in the test media was simulated through combined functions of the DATA- GEN4 software, which compiles all data information, and the WQ4F software, which calculates the activities (and/or concentra- tions) of major ions and ion pairs; programs were written by the U.S. Geological Survey in 1988. For larval swimming tests, each well on the culture plates contained a total of 1.500 (jlL. obtained through the sequential pipette additions of 1 .000 \xL of seawater. 350 jj-L of seawater containing 20-30 larvae, and 150 [xL of 10 x concentrated metal solution prepared in distilled water. For the control, 150 jxL of distilled water or seawater was added in place of metal solution. Larval swimming behavior was observed under a dissecting mi- croscope during the 96-h trial . Larval responses to Cu and Zn were compared with the control with respect to the speed of swimming, ciliary activity, and the number of quiescent or inactive larvae. Larval searching behavior was induced by 10^"* M l-3,4- dihydroxyphenylalanine (l-DOPA) (Coon et al. 1986). Induced larvae begin swimming with the foot extended forward, then sink to the bottom, withdraw the velum, thrust the foot anteriorly, and begin crawling. Larvae, in this stage, may resume swimming if the substratum is in some way inappropriate for cementation. Larvae were preexposed (preacclimated) to metal-amended seawater for 96 h before being induced by L-DOPA (10 "* M). Behavior was monitored every 5 min for 30 min, and the fraction of searching larvae in each well was recorded. Oyster larval metamorphosis requires suitable biofilms in na- ture but can be artificially induced by exposing competent larvae to 10"^ M EPI in seawater for 4-12 h (Coon et al. 1986). Larvae respond by immediately sinking to the bottom, without searching behavior, progressively resorbing the velum, and eventually show- ing significant new shell growth in 24-48 h. Experiments were designed to differentiate metal inhibition periods for larvae ex- posed to metals: (A) during induction with EPI, (B) after induction with EPI, and (C) during and after induction with EPI. In the first interaction, larvae were exposed to metals with EPI for 4 and/or 12 h before the medium was replaced by clean seawater, in which larval metamorphosis was scored after 96 h. In the second inter- action, larvae were exposed to EPI without adding metals; the medium was then replaced by metal-amended seawater. In the third interaction, larvae were exposed to EPI with metals and then were continuously exposed to metal-amended seawater for 96 h and larval metamorphosis was examined. Larval Set on Biofilms We used Hyphomonas because it is a ubiquitous marine bac- terium with a role in biofilm community ecology (viz. invertebrate set (Quuitero and Wciner 1995). Bacterial biofilms of Hvplwmo- nas spp. strain PM-I were generated on siliconized glass beakers (Sigmacote, Sigma Chemical Co., St. Louis. MO) as previously described (Chang 1995). All filmed beakers were rinsed first with clean MBL seawater and were individually filled with 50 mL of MBL seawater (salinity. 37 g ■ kg"') ; then. 100 or so competent C. gigas larvae were introduced through a micropipette. Beakers were then rotated gently at 4 rpm for 96 h at 30°C. The water level was maintained at the 50-mL mark by dripping in distilled water on a daily basis. At the end of the incubation period, vessels were examined individually for set (cemented and/or metamorphosed) larvae un- der a dissecting microscope. Larvae that did not settle on the surface were collected by Nitex screens (170 |jim pore size) and were counted. The total number of larvae applied to each individ- ual beaker was the sum of set and unset larvae. Effects of Biofilms on Metal Toxicity For metal toxicity tests, bacterial filmed surfaces were exposed to Cu or Zn and then were presented to competent larvae. Filmed beakers produced from the bacterial cultures first were rinsed by MBL seawater and filled with 50 mL of MBL seawater (salinity. 37 g ■ kg" ' for C. gigas and 24.7 g • kg " ' for C. virginica): Cu or Zn were then spiked to make final total metal concentrations of 0. 100. 500, and 1 ,500 ppb. After a period of 24 h, during which metals were exposed to the biofilms (Chang 1995), competent larvae were introduced; larvae that cemented and metamorphosed on filmed surfaces were counted after 96 h. For individual metal treatments, five replicates of filmed beakers were prepared. BiOFiLMS Magnify Cu and Zn Toxicity to Oyster Set 591 Chesapeake Bay Samples Water samples were collected from two sites in the Chesapeake Bay. One site was at the northern bay near the mouth of the Patapsco River in the Baltimore Harbor area (an area of heavy industry), and the other was in the middle bay (Tilghman Island and Chesapeake Beach) (an area that is relatively pristine). Acid- washed and autoclaved sample containers (10 L polyethylene) were totally immersed until fully filled. This estuarine water was stored below 4°C and transported back to the laboratory for C. virginica or C. gigas larval swimming, searching, metamorphosis. and .settling experiments and for metal analyses. For larval settling experiments, laboratory-produced biofilms were used as the settling surfaces. Collected bay water (50 mL) was then added to individual filmed beakers (five replicates for each water sample). C. gigas larvae were introduced, exposed, and examined as described above Most bay water contained Cu or Zn below the detection limits of tlame AAS; therefore, water samples first were digested to release bound metals from organics and then were evaporated to reduce volume to one-tenth of the original. These were then five-fold-concentrated by chelation with ammonium pyrrolidine dithiocarbamate and extraction into methyl isobutyl ketone (Am. Pub. Hlth. Assoc. 1990). Water salinity at the collection sites, measured with a refractometer, was approxi- mately 5-10 g • kg^ ' at Baltimore Harbor and 1 1-18 g • kg ' at the middle bay. RESULTS Metal Toxicity Tests The effects of Cu and Zn on larval search behavior and devel- opment are shown in Table 1 . C. gigas larval swimming was not inhibited by total Cu concentrations up to 1,500 ppb (i.e., 24 ppb Cu" * as calculated with WQ4F software) or total Zn concentration up to 500 ppb (i.e., 200 ppb Zn" ^ ). However, larval swimming was significantly reduced by 1,5(X) ppb Zn (i.e., 610 ppb Zn"*) (Mest, p < 0.01). C. gigas larval searching behavior (induced by L-DOPA) was inhibited by total Cu at 1,500 ppb and more so (p < 0.01) by total Zn at 1,500 ppb. C. gigas metamorphosis, arti- ficially induced EPl, was inhibited (p < 0.01) by 1,500 ppb total Cu and by total Zn &500 ppb. As total Zn concentration ap- proached 1,500 ppb. all larvae remained quiescent. C. virginica larvae were similarly but not identically affected by Cu and Zn (Table I ). The data were influenced by the fact that the ratios of free ion to total metal concentrations of Cu and Zn TABLE 1. Effects of copper and zinc on oyster larvae, C. gigas and C. virginica, including swimming activity, searching behavior, metamorphosis, and set on biofilm. Oyster (Salinity) Metal iivir (ppb) |M-*]' (ppb) Swim' Search" Metamorphose' -1- -1- + + + + + + + + + + + + + + + + + + + + Set on biofilm' C. gigas (37g-kg" C. virginica (24.7 g ■ kg- Cu Zn Cu Zn 0 0 + +^ + + 100 1.6 + + + + 500 8.0 + + + + 1.500 24.0 + + + + 0 0 + + + + 100 40 + + + + 500 200 + + + + 1 ,500 610 + - 0 0 + + + + 100 1.7 + + + + 500 8.3 + + + 1.500 25.0 + ' - 0 0 + + -1- -1- 100 50 + + -F -1- 500 250 + + + 1 .500 750 + + + + + + + + + + + + + + + + ' Initial dissolved absolute metal concentration in seawater '' Estimated concentrations of free Cu"* or Zn"* ions according to metal speciation program (WQ4F) at pH 8 in specific salinity of MBL seawater. ' Swimming activity after 96 h. " L-DOPA-induced searching behavior after 96 h of metal exposure. ' EPI-induced metamorphosis. Metals were added with EPl for 4 h of exposure; then, the water containing EPI was removed and replaced by water containing only appropriate concentration of metals. ' Biotllms of 5-day cultures. Marine broth was replaced with metal-amended MBL seawater 24 h (penod of metal-film bmding reaction) before the addition of competent oyster larvae. ^ Activity response: -I- -I- . >80% vs. control; + . 35-80'J vs. control; - . <35% vs. control. "No" metal controls were done with every individual experiment, so valid comparisons could be made. Control responses varied with larval batch and other variables as follows; Swim, 90-100%; Search. 70-100%; Metamorphose. 50-90%; Film set. 20-70%. '' 83% vs. control but statistically significant (p < 0.01). ' 75% vs. control but statistically insignificant at a = 0.01. 592 Chang et al. were, respectively. 4 and 23% higher in 24.7 g of salts • kg"' seawater (C. virginica) than in 37 g of salts • kg"' seawater (C. gigas). Swimming and search behavior were affected more by Cu and less by Zn than for C. gigas. On the other hand, artificially EPI-induced metamorphosis was affected approximately equally by both metals in both species of Cnissostrea. The last column of data, set on biofilms, should be assessed considering at least three additional components that influence the results; (A) set on biofilms must be preceded by search behavior (see introduction) and cementation before metamorphosis, so that the additional factors could make the process more sensitive to potential toxics; (B) the metals are bioconcentrated to some extent; (C) the metals in the biofilms may not be totally available (see Discussion). The net result was that, in the presence of biofilms. the exper- imental aqueous metal concentrations became more inhibitory to metamorphosis (Table 1 ). To compare one treatment with another, the effective concentrations of metals that cause 50% reduction in larval normal responses (EC^o) were estimated. The aquatic EC^,, of initial total dissolved Cu or Zn for EPI-induced larval meta- morphosis of C. gigas decreased from 1 .200 ppb Cu or 500 ppb Zn to < 100 ppb Cu or <200 ppb Zn for natural set in the presence of biofilms; that for C. virginica decreased from 1 ,000 ppb Cu and 330 ppb Zn to 500 ppb Cu and 170 ppb Zn, respectively. There- fore, in the presence of biofilms. the detrimental effects of Cu increased more than 10-fold for C. gigas and twofold for C. vir- ginica, and those for Zn increased threefold and twofold, respec- tively. In situ, larvae would be continuously exposed to environmental concentrations of Cu and Zn. However, the following series of experiments asked the question of whether the toxic effects are exerted during or after the induction signal. Both Cu and Zn ex- erted their most profound effects after mctamorphic induction with EPI (Fig. 1). There was significant inhibition in postinduction metamorphic development at 1 .500 ppb total dissolved Cu and &500 ppb total dissolved Zn: inhibited larvae remained quiescent. Larval metamorphosis was usually not affected when metals were present only during the EPI-induction period, in the presence of 1.500 ppb Cu, a longer EPI-induction period (12 h) was required than in controls; those larvae that did not metamorphose by 4-h EPI induction remained swimming in the water column. Thus, induction time and metal concentration both infiuence EPI- induced metamorphosis. As was the case for C. gigas, the EPI-induced metamorphosis of C. virginica was inhibited more by coincident exposure to Cu than to Zn (s500 ppb total dissolved Cu and ss 1.500 ppb total dissolved Zn) (Fig. 2). The postinduction metamorphic develop- ment of C. virginica. like that of C. gigas. also was inhibited by 1,500 ppb Cu (i.e., 25 ppb Cu"*) and &500 ppb Zn (i.e., s250 ppb Zn-* ). Only a 4-h EPI exposure period was used, because at longer induction periods, C. virginica larvae responded errati- cally. Chesapeake Bay Samples Because experimental models suggested that biofilm bacteria concentrated metals to levels where metamorphosis was inhibited. it remained to use the biofilm model (bioassay) to test natural Chesapeake Bay water, which contains very low concentrations of these metals. Even when concentrated 50-fold by the APDC/ 125 100 c o o x: Q. O E IT! TO £ 25 0 125 100 Cu 75 - 50 - 25 - [Metal] (ppb) Figure 1. Effects of timed exposure to Cu and Zn on C, gigas larval metamorphosis (larval response normalized to control). Experiments were divided into three parts: exposure to EPI with metals for 4 (U) or 12 h (D) and then to clean seawater for 96 h; exposure to EPI without metals for 4 (A) or 12 h (V) and then to metals for 96 h; exposure to EPI with metals for 4 ( : ) or 12 h ( O ) and then to metals for 96 h. Error bar = 95% confidence interval. MIBK extraction method, total Cu and Zn in water near the Bal- timore Harbor and middle bay remained undetected by the meth- ods used here. Because 50 ppb was the preconcentration detection limit, it would mean that even in polluted Baltimore Harbor, the concentration was <1 ppb. Nevertheless, total percent larvae ce- mented on control biofilms was less in Baltimore Harbor water than in middle bay water or MBL seawater (Table 2). Levels of metamorphosis of set larvae were at least as high in harbor and bay water compared with artificial seawater. This finding makes use of the system as a bioassay. and any one of a number of factors, together or in combination, could have influenced set. Salinity was not an overriding variable, however, because harbor water sup- ported more metamorphosis than did artificial water. Signifi- cantly, harbor water did not inhibit C. virginica larval swimming, searching, and EPI-induced metamorphosis; it did inhibit biofilm- induced set in the model bioassay organism, C. gigas. BiOFiLMS Magnify Cu and Zn Toxicity to Oyster Set 593 200 150 - Cu c o o o m ■(0 o CL i_ o E 3 CO 150 100 50-1 500 1000 1500 [Metal] (ppb) Figure 2. Effects of timed exposure to Cu and Zn on C. virginica larval metamorphosis (larval response normalized to control): C , ex- posure to EPI v»ith metals for 4 h and then to clean seawater for 96 h; D, exposure to EPI without metals for 4 h and then to metals for 96 h: and A, exposure to EPI with metals for 4 h and then to metals for 96 h. Error bar = 95% confidence interval. DISCUSSION Metal speciation in seawater affects bioavailability. Experi- ments were carried out in environmentally relevant salinities. In MBL seawater with salmity 24.7-37 g • kg" ' at pH 8, free avail- able Zn^"^ ions comprised about 40-50% of the total Zn, but free available Cu~"^ ions comprised only about 2% of the total Cu. Thus, when total metal concentrations are considered. Zn is more toxic than Cu. However, when concentrations of free Cu"* and Zn"* are compared (Table 1), the opposite is true. Normally, the length of the EPI-induced period required for competent larvae to metamorphose is 2 h for >80% metamorpho- sis (Coon et al. 1986). However, a longer period of EPI exposure was required to induce C. gigas larvae to metamorphose in 1 ,500 ppb Cu-amended seawater (37 g • kg" '). Cu may have disturbed EPI induction through its interference with larval EPI receptors, or through its chemical complexation with EPI. Zn coordinates fewer TABLE 2. Effects of Chesapeake Bay water and MBL seawater on C. gigas larval set on biofilms. Water" Salinity (g g ' % Attached'' % Metamorphosed' MBL MCB EH 37 12 6 63 ± 7 53 ± 15 35 ± 27 18 ± 7 39 ± 20 26 ± 20 ■* MBL, Marine Biological Laboratory artificial seawater; MCB, middle Chesapeake Bay water; BH, Baltimore Harbor water. '' Percentage (mean ± 95% confidence interval) of total cemented larvae/ total larvae. •^ Percentage (mean ± 95% confidence interval) of metamorphosed ce- mented larvae/total larvae. ligands and only inhibited the EPI induction of C. virginica in seawater with lower salinity (24.7 g ■ kg" '). Most significantly. Cu or Zn concentrations of 100 ppb in estuarine water did not affect oyster larval viability, swimming, DOPA-induced searching behavior, or EPI-induced metamorpho- sis; however, they unequivocally inhibited larval set in the pres- ence of biofilms compared with control films, which had not been exposed to additions of Cu or Zn. Although Cu was less toxic than Zn in seawater because less Cu"* than Zn^* was present, Cu became more toxic to larval set on biofilms, implying that Cu was more bioavailable than Zn in the films. Biofilm Cu concentrations at around 15-20 ppm ((xg/g of biofilm wet wt.) measured directly in biofilms prevented 50% C. virginica larvae from set (Chang 1985). Aqueous concentrations of dissolved total Cu that might lead to this bioconcentration were estimated to be 0.1-0.2 ppm. Phelps and Mihursky (1986) also found that competent larval oysters (C. virginica) had decreased set with increasing Cu con- centrations. However, they found an LDjo of 534,000 ppb, con- siderably higher than that reported here and elsewhere (e.g., Mit- tlcman and Gcesey 1985, Calabrese et al. 1973). It is possible that the aufwuchs did not readily release detectable metal species in that particular system. Bound metals can be released from biofilms through highly dynamic biologic activities (e.g., redox and pH gradients in bio- films) (Bender et al. 1995). Homor (1984) indicated that aquatic heterotrophs can leach and solubilize metallic sulfides at neutral to slightly acidic pH and release heavy metals to the water column under oxygen-limited conditions. The author concluded that two mechanisms were responsible for the leaching; (A) the microbial production of solubilizing agents (i.e., low-molecular-weighl or- ganic chelating agent) that diffuse up from sediments and compete with anionic ligands (sulfides) for metallic complex formation: and (B) sulfide oxidation by facultative anaerobes or microaerophiles. Francis ( 1990) also summarized the mechanisms of metal mobili- zation mediated by heterotrophic microbes in mixed wastes. These include (A) changes in pH that determine solubility; (B) oxidation- reduction reactions and redox potential, which affect valence states and solubility; (C) chelation, solubilization, and leaching of elements by microbial metabolites and decomposition products; (D) biomethylation and production of volatile and/or toxic alkyl- ated metal compounds; (E) biodegradation of organic complexes of metals; and (F) replacement of Ca"* by heavy metals from anionic ligands (Snoeyink and Jenkins 1980). Biofilm-bound met- 594 Chang et al. als could also be released through the ingestion of EPS by the low pH in the gut. Metal-contaminated biofilms. fed upon by larvae, would be a route of entry, because some bacteria serve as food sources or growth factors (Douillet and Langdon 1993. Zobell and Allen 1935). The 12-day LC50 of Cu for 2-day-old C. virginica D-hinged larvae was reported at 32.8 ppb (Calabrese et al. 1977). Results reported here showed that larvae retained swimming activities (70% of the control) at 1,500 ppb Cu for 4 days. In addition to time of exposure, variations may be attributed to the earlier study's selection of D-hinged larvae versus older competent larvae here; younger larvae generally are more susceptible to pollutants than older larvae. Calabrese et al. (1977) reported that the pH was maintained at 7.0-8.5. which would vary [Cu""^] from 30% to <2% of the total [Cu]; this could result in toxicity variations >15 times. In this study, the pH was maintained around 8.0. at which [Cu"^] comprised about 2% of the total Cu. Last, m the earlier study, larvae were continually fed with algae during the test pe- riod; therefore, metal contamination and bioconcentration through algae become a factor. Each of these variables would lower the LC5,). making the results more comparable with this report. It is significant that these bioassay systems are very sensitive to spe- cific habitat conditions (Phelps and Mihurksy 1986. Calabrese et al. 1973. Calabrese et al. 1977. this report), underscoring their potential utility in assessing toxicity in situ. The data with our biofilm/oyster system, as a bioassay for Chesapeake Bay water, were interesting. Seven to 15 ppb Cu""^ and 150-230 ppb Zn"* inhibited EPl-induced Crassostrea meta- morphosis. Moderately polluted portions of the Chesapeake Bay have been shown to contain 1.6 x lO""* to 5 x 10~' ppb Cu"^* and 3-50 ppb Zn" * ; less polluted portions ha ve 1 . 6 x 1 0 '* to 3 . 2 X I0~^ ppb Cu-* and 0.1-0.5 ppb Zn"* (Sunda et al. 1990). Preliminary data of bioconcentration factors of autochthonous films (i.e.. ratio of metal concentration in films over that in am- bient water) for these metals ranged from 6.000 to 8.000 for Cu and 1.000 to 2.000 for Zn (biofilm dry wt.), depending on the physical parameter and the type of biofilm (Chang 1995). Thus, the predicted concentration in biofilms in 0.96-40 ppb Cu"^ * and 3 X lO"* to 1 X 10^ ppb Zn"* in moderately polluted areas and 0.1-0.3 ppb Cu-* and 100-1.000 ppb Zn"* in less polluted ar- eas. If these metals were made available to the larvae, as our biofilm studies suggest, then Cu, in moderately polluted portions of the bay, and Zn in both the polluted and the less polluted portions of the bay, would be concentrated enough to affect set. The results are consistent with that suggestion. The concentration of Cu in sediments of the Chesapeake Bay may range from < 2.0 to 130 ppm and that of Zn may range from <10 to 600 ppm (Wright 1987). If even 0.005-0.8% of Cu or 0.03-2% of Zn partitioned out of the sediments and into set- inducing biofilms, it could affect oyster recruitment. In any case, this study clearly demonstrated that metal toxicity to larval set can be increased by 10- fold in the presence of biofilms. Metamorpho- sis from the metalarva to the juvenile stage often is used as a sensitive indicator in toxicity tests (Phelps and Warner 1990). because during the transition, there is considerable anatomical reorganization, the larvae are under stress, and toxic ions are more likely to interfere with their metabolic processes. All of this occurs on biofilms (Prieur et al. 1990). Larvae that fail to metamorphose occurs on biofilms (Prieur et al. 1990). Larvae that fail to meta- morphose die. Thus, toxic metals may be bioconcentrated where and when oysters are most sensitive, which should be considered in pollutant regulations. ACKNOWLEDGMENTS We thank G. Helz for guidance on metal chemistry, B. Schafer for statistical analysis, and J. Kovach for critical reading of the manuscript. This project was supported in part by grants from the Maryland Water Resources Research Center. University of Mary- land, and #NA90-AAD-000I4 from the Maryland Sea Grant Col- lege. A portion of this work was submitted in partial fulfillment of the requirements for the Ph.D. degree for E. Chang in the Marine and Estuarine Science Program at the University of Maryland, College Park, MD (Chang 1995), and a preliminary report was presented at the 93rd Annual Meeting of the American Society for Microbiology (Chang et al. 1993). LITERATURE CITED Amencan Public Health Association. American Water Works Association, Water Environment Federation. 1990. Standard Methods for the ex- amination of Water and Wastewater. 17th ed, American I\iblic Health Association. Washington. DC. Bender. J.. R. F. Lee & P. Philllp^. 1995. Uptake and Iransfomiation of materials and metalloids by microbial mats and their use in hioreme- diation. J. Indusl. Microbiol. 14:113-118. Bonar. D. B., S. L. Coon. M. Walch. R. M. Weiner& W. K. Fitt. 1990 Control of oyster settlement and metamorphosis by endogenous and exogenous chemical cues. Bull. Mar. Sci. 46:484—498. Boyle. E. A.. F. R. Sclater & J. M. Edmond. 1977. The distnbution of dissolved copper in the Pacific. Earth Plane! Sci. Lett. 37:38-54. Calabrese. A.. R. S. Collier. D. A. Nelson & J. R. Maclnnes. 1973. The toxicity of heavy metals to embryos of the American oyster Crassos- trea virginica. Mar. Biol. 18:162-166. Calabrese, A.. J. R. Maclnnes. D. A. Nelson and J. E. Miller. 1977. Survival and growth of bivalve larvae under heavy-metal stress. Mar. Biol. 41:1779-184. Chang. E. 1995. Impact of bacterial biofilms in the toxicity of Cu and Zn to oyster larvae. Ph.D. dissertation. University of Maryland Library. College Park. MD Chang, E.. R. Weiner & M. Walch. 1993. Impact of bacterial biofilms on the toxicity of copper and zinc to oyster development. Abst. Gen. Meet. ASM Atl. 93:20. Coon, S. L., D. B. Bonar & R. M. Weiner. 1985. Induction of settlement and metamorphosis of the Pacific oyster, Crassostrea gigas (Thun- berg), by L-DOPA and catecholamines. J. E.xp. Mar. Biol. Ecol. 94: 211-221. Coon. S. L. D. B.. Bonar & R. M Weiner. 1986. Chemical production of cultchless oyster spat using epinephnne and norepinephrine. Aqiiaciil- ture 58:255-262. Coon. S. L.. M. Walch. W. K Fitt. R. M. Weiner & D. B. Bonar 1990. Ammonia induces settlement behavior in oyster larvae. Biol. Bull. 179:297-303. Crisp, D. J. 1974. Factors influencing the settlement of marine inverte- brate larvae, pp. 177-265. In: P. T. Grant and A. M. Mackie (eds.). Chemoreception in Marine Organisms. Academic Press. New York. Douillet, P. & C. J. Langdon. 1993. Effects of manne bacteria on the BiOFiLMS Magnify Cu and Zn Toxicity to Oyster Set 595 culture of axenic oysler Crassoslrea gigas (Thunberg) larvae. Biol. Bull. 184:36-51 Fitt, W. K. & S. L. Coon. 1992. Evidence for ammonia as a natural cue for recruitment of oyster larvae to oyster beds in a Georgia salt marsh. Biol. Bull. 182:401^08. Fitt, W. K.. M. P. Labare, W. C. Fuqua. M. Walch. S. L. Coon. D B Bonar. R R. Colwell & R. M. Weiner. 1989. Factors intluencmg bacterial production of inducers of settlement behavior of larvae of the oystST Crassoslrea ,?i,i; Yield (mL) (mLl (%) (%) (%( 1%) (%) (XIO ■*) Comitan Suspended Bottom Bahia Falsa Suspended Bottom 50 65 7 52 1780 0,342 91 7 44 1 17 22 49.88 9.08 37.79 3.25 4.0 100 65 0 51 17.28 0.339 91.8 41.4 18.77 50.01 8.62 38.19 3.18 5.0 150 65.1 47 15 24 0.324 90.8 38.9 18.63 49,97 8.69 38.11 3.23 3.0 50 73.2 59 20.37 0.345 87.8 40.1 17.61 51.33 7.93 37.53 3.21 3.0 100 73.1 55 IP 34 0,352 87,9 39.8 17.89 51.02 8.09 37.70 3.19 3.0 150 72 7 50 16.85 0.337 87.7 36.3 18.09 51 12 8.10 37.56 3.22 3.0 50 71 5 64 23.69 0.370 92.5 38 0 22.88 49.34 8.04 39.27 3.35 1.0 100 71.4 52 17 81 0.343 90.1 36.2 2097 50.17 8.13 38.47 3.23 2.0 150 72.1 48 13.16 0274 920 36 3 21.81 50.19 8.07 38.65 3.09 1.0 50 78.5 72 27.25 0 378 78,7 262 21.34 53.01 8.01 35.72 3.26 1.0 100 78.3 57 20.81 0 365 69 9 25,4 21 09 52 24 7.69 36.80 3.27 1.0 150 73 3 50 16 85 0 337 73 4 27 3 21 21 52 57 7,80 .36.72 3.21 1.0 in the warmer months of summer (August and September) and in the first year of life. The greatest production (Table 5) was in July of the first year, with a value of 6.61 g per individual for total. Even though this was the highest monthly production value, the greatest yearly production was at the second area (Bahia Falsa) under suspended conditions and at a density of 50 organisms/m". In all treatments, production decreases from December 1987 to January 1988, in many cases resulting in negative values. A dif- ference between production and biomass is noticeable at this point, because biomass levels vary little compared with production lev- els. After this production decrease, it began to increase again, but less than in the first year. Production was close to 0 between August and September (summer season) of year 2. The multifac- torial (ANOVA) variance analysis test was applied to determine if differences exist in annual somatic production for each of the treatments (Table 7). The means of annual oyster production are different for the areas tested (p < 0.0003), with the greatest values at site 2 (BF). There are also greater statistical differences in culture system and density, with the largest values at the bottom and at a density of 50 organisms/m". DISCUSSION Comparing growth and mortality results obtained for C. gigas. Islas et al. (1982) give a monthly length-growth rate for the Jap- anese oyster of 9.2 mm and mortality rates of 20% over 6 mo for a site on the west coast of Baja California (approximately 28°N). At the same latitude, but in the Gulf of California. Ochoa and Fimbres (1984) found an average monthly length and weight growth of 7 mm and 7.7 g with a mortality of 30% over 10 mo. In Bahia de La Paz. I found slower growth in length. 6 mm/mo the first 6 mo. and at 10 mo. 5.3 mm/mo for length. 3.6 g/mo for weight, and greater mortality (35%). Acosta (1985) obtained av- erage growth of 10 mm' mo and 6 g/mo over a year at an area at 30°N along the west coast of Baja California. In Japan. Fujiya ( 1970) and Korringa ( 1976) report growth for the first year of 60 g. Askew ( 1972) and Hall ( 1984) in Great Britain cite total weights of 55 and 70 g at 12 and 16 mo. respectively. Berthome et al. (1986) obtained total annual growth of 40 g at 46°N along the Atlantic Coast of France. In the Bay of Arcachon, Robert et al. (1993) achieved an annual mean weight of 35 g. The results of Berthome et al. are similar to the 43 g I obtained in a suspended system at site I with 50 organisms/m" (CO50). less than the 55 g in the BF50 treatment at the bottom, site 2 with 50 organiams/m~, and the same as Shafee and Sabatie (1986) for an area at 36°N on the coast of Morocco. When monitoring proteins, lipids, and carbohydrates as mea- sures of production quality, as many authors call them (Baird 1957, Reinitz and Yu 1981, Boggio et al. 1985, Bodoy et al. 1986, Viola et al. 1988), my objective was not to study changes over time but rather to measure proteins, lipids, and carbohydrates at the end of the study as a comparison of changes in nutritional components over a year. Paez et al. (1993) did this with two species of Crassostrea in a study near Bahia de La Paz. The results of the analysis of the relationship between the ex- TABLE 5. Probability levels and statistic F obtained from values of meat yield, condition factor, protein, lipids, and carbohydrates by the use of a multifactorial ANOVA on the factors: site, system, density," ANOVA Meat Yield Condition Factor Proteins Lipid Is Carbol F dydrates Factors F P F P F P F P P Site System Density 1.307 0.461 0.021 0.2753 0.5170 0.9788 0.963 0.859 0.803 0.3533 0.3794 0.4675 12.76 9.773 0.631 0.0031 0.0074 0.5466 0.261 0.839 0.185 0.6227 0.3849 0.8333 3.492 2.572 0 108 0.0827 0.1311 0.8985 ' No interactions were significant. Production. Growth, and Survival of C. Gigas 605 TABLE 6. Production matrix of cohort in site I (CO), on the suspended system culture (C), and with a density SO/m^." L WW WI DMW DI DMB Month (mm) (g) N (gt WWP (g) (g) DM? (g) P/B 1987 M 10.0 0.03 1 .000 0.003 2.5 A 24.4 3 1.000 2.97 2,965.0 0.21 0.21 207.6 210 0.99 M 25.8 7 944 3.777.8 0.49 0.28 264.4 462.8 0.57 J 26.7 12 944 4,722.2 0.84 0.35 330.6 793.3 0.42 J 30.1 19 944 6,611.1 1.33 0.49 462.8 1.256.1 0.37 A 32.2 23 944 3,777.8 1.61 0.28 264.4 1,520.6 0.17 S 33.3 26 944 2,833.3 1.82 0.21 198.3 1,718.9 0.12 O 34.9 29 833 2.500.0 2.03 0.21 175.0 1.691.7 0.10 N 41.7 33 722 2.888,9 2.31 0.28 202.2 1.668.3 0.12 D 56.1 39 667 4.000.0 2.73 0.42 280.0 1,820.0 0.15 1988 J 61.0 39 667 0.0 2.73 0.00 0.0 1,820.0 0.00 F 63.3 41 611 1.222.2 2.87 0.14 85.6 1,753.9 0.05 M 63.9 43 444 888.9 3.01 0.14 62.2 1,337.8 0.05 A 64.4 45 444 888.9 3.15 0.14 62.2 1,400.0 0.04 M 64.8 46 444 444.4 3.22 0.07 31.1 1,431.1 0.02 J 65.1 47 389 388.9 3.29 0.07 27.2 1,279.4 0.02 J 65.3 48 389 388.9 3.36 0.07 27.2 1,306.7 0.02 A 65.5 49 389 388.9 3.43 0.07 27.2 1,333.9 0.02 S 65.6 49 389 0 0.0 3.43 0.00 0.0 1,333.9 0.00 O 65.6 50 389 1 388.9 3.5 0.07 27.2 1,361.1 0.02 N 65.7 52 389 2 777.8 3.64 0.14 54.4 1.415.6 0.04 First-year prod luction per 20 m- 36.19 kg 2.53 kg First-year monthly production per m" 3,015.6 g 211.1 g Total production per 20 m" 39.85 kg 2.79 kg Total monlhly production per m" 1.992.7 139.5 g " Units recorded for standardizing to 1,000 organisms (N) were: length (L), individual wet weight (WW), monthly wet weight increment (WI), dry meat weight (DMW) dry rneat weight increment (DI). dry meat biomass (DMB), and dry weight produclion/biomass quotient (PB). Wet weight production (WWP) and dry meat production (DMP) were calculated in grams per square meter by month perimental factors affecting growth (in length and weight), mor- tality, condition state, nutrition, and secondary production were analyzed with a multifactorial ANOVA to observe if differences existed among the conclusions made by considering the parame- ters separately and for decision making in aquaculture (Table 8). Higher values of growth rate and mortality were obtained at site I (CO). At site 2, high production was found. It is difficult to account for site differences because there are few studies of the region. Annual water temperature and salinity averages are higher at the inner site (25.8°C and 36.2 salinity) than at site 2 (24.7°C and salinity of 35.0). Lango (1994) showed the annual mean of other water variables for sites 1 and 2 to be: 5.2 and 5.6 mg/L of dissolved oxygen, 0.47 and 0.44 mg/L of seston, 0.40 and 0.037 mg/L of plankton, and 0.006 and 0.005 mg/L of tripton. There is nothing recorded about hydrodynamic conditions. Depth and sed- iment composition (fine sand) are also similar at the two sites. However, site 2 had a higher secondary production than site I . At site 1 (CO), predators of bivalves like the fish Sphoeroides annu- laius (Jenyns), snails of the genus Strombus, and the flatworm Figure 2. Curves of dry meat biomass (DMB), wet weight production (WWP), and dry meat production (DMP) at site 1, on the suspended culture system, at a density of 50 organisms/m^ (Exp. code COC50). TABLE 7. Probability levels and statistic F obtained from values of production by the use of a multifactorial ANOVA on the factors: site, system, and density, and their interactions. ANOVA— Production Factors Site System Density Interactions Site-system Site-density System-density 23.042 64.578 21.980 0.339 1.567 4.799 0.0003 0.0000 0.0000 0.5759 0.2431 0.0259 606 Arizpe TABLE 8. Overall results from multifactorial ANOVA of cohorts in C. gigas as a function of meat yield, condition factor, biochemical variables, K of Bertalanffy, weight growth rate, mortality, and the variables of somatical secondary production related to the experimental factors and their interactions." Parameter Z Site S System D Density Interactions ZxS ZxD SxD Meat yield Condition factor Proteins Lipids Carbohydrates Bertalanffy K Weight growth rate G Morathty rate Z Production Meat yield Condition factor Proteins Lipids Carbohydrates Bertalanffy K Weight growth rate G Mortality rate Z Production BF CO CO BF F C 50 ' The specific factors from which the largest values were obtained are shown in the lower half. [( - ) p > 0.05, (*) p « 0.05. (**) p =5 0.01] Srylochiis sp. . are common. This caused higher mortality at site 1 . not just for C. gigas but also for a native bivalve. Pinna riigosa (Arizpe 1995). Oysters are better protected against predation in the suspended trays and have lower mortality. Contrary to tropical bivalves like P. rugosa, local C. gigas production decreases in summer. The surface water temperature reaches 28-29°C in September. Statistical evidence exists for a density effect on production only (Table 8). The density effect did not appear in growth parameters, mortality, biochemical compo- sition, meat yield, or condition factor. Mortality of growth rate, taken separately, can give contrary results, as in the case of site 1 (CO), which had better weight growth rate (G) but also high mor- tality. Using only one criterion such as mortality or growth in- creases the difficulty for a decision with respect to site location. Production showed a higher sensitivity to site location, culture system, and density, than to the other factors, perhaps because its determination includes both mortality and body growth. Maximiz- ing production is directly relevant to the success of aquaculture ventures. Small-scale plantings of similar-sized, hatchery-reared spat and the monitoring of population production are the best approaches for assessing site suitability for the culture of C. gigas. ACKNOWLEDGMENTS I am indebted to Dr. David H. Cushing of Lowestoft Research Laboratory, England, for his encouragement and advice during this study. Thanks also to Dr. Ellis Glazier. CIBNOR, for his critical editing of the English language manuscript and to Julio Garcia for data processing. LITERATURE CITED Acosta. M. R. 1985. Eficiencia nutricional del ostion japones. Crasso.s- trea gigas (Thunberg) en Bahi'a San Quintin e Isia San Martin, B.C.. Mexico. Tesis de Maeslria. CICESE. Mexico, Allen, K. R. 1951. The Horokiwi Stream; a study of trout population Fish. Bull. N.Z. 10:1-238. A.O.A.C. 1980. Official methods of analysis. 13th ed. Association of Official Analytical Chemist. Washington, DC. Arizpe, O. 1987. Reclutamiento y mortalidad de Pinna rugosa en condi- ciones semicontroladas en Bahia de La Paz, Mexico. Anal. Inst. Cien- cias Mar Limnol. UNAM 14:249-254. Arizpe, O. 1995. Mortality, growth and somatic secondary production of the bivalve. Pinna rugosa (sowerby), in suspended and bottom culture in Bahia de La Paz, Mexico. Aquacult. Res. 26:843-853. Arizpe, O. & R. Felix. 1986. Crecimiento de Pinna rugosa (Sowerby. 1835), en Bahia de La Paz, Mexico. Anal. Insl. Ciencias Mar Limnol. UNAM 13:167-172. Askew, C. 1972. The growth of oysters Ostrea edulis and Crassostrea gigas in Emsworth Harbour. Aquaculture 1:237-259. Baird. R. H. 1957. Measurements of condition in mussels and oysters. J. Conseil. Denmark 23:249-257. Baqueiro, E. & M. Castagna. 1988. Fishery and culture of selected bi- valves in Mexico: past, present and future. J . Shellfish Res. 7:433-443. Begon, M. & M. Mortimer 1986. Population Ecology. Blackwell Scienlic Publishers, Oxford. Berg. C. & D. Olsen. 1989. Conservation and management of queen conch. (Strombus gigas) fisheries in Caribbean, pp. 421—442. In: J. F. Caddy (ed.). Marine Invertebrate Fisheries: Their Assessment and Management. John Wiley. New York. Berthome, J., J. Prou & A. Bodoy. 1986. Performances de croissance de Phuitre creuse. Crassostrea gigas (Thunberg) dans le bassin d'elevage de Marennes-Oleron entre 1979 y 1982. Halioiis 15:183-191. Beukema. J. & W. De Bruin. 1987. Seasonal changes in dry weight and chemical composition of the soft parts of the Tellinid bivalve, Macoma balthica in the Dutch Wadden Sea. Neth. J. Sea Resources 11:42-55. Beverton, R. J. H. & S. J. Holt. 1957. On the dynamics of exploited fish populations. M.A.F.F. (London). Fish. Invest. Ser. 2(19). Bodoy. A.. J. Prou & J. Berthome. 1986. Etude comparative de differents indices de condition chez I'huitre creuse Crassostrea gigas. Haliotis 15:173-182. Boggio, S.. R. Hardy. J Babbitt & E. Brannon 1985. The influence of dietary lipid source and alpha-tocopheryl acetate level on product qual- ity of rainbow trout (Salmo gairdneri). Aquaculture 51:13-24. Brown, J R. & E. B. Hartwick. 1988. A habitat suitability index model for suspended tray culture of the Pacific oyster. Crassostrea gigas Thunberg. .Aquacult. Fish. Mgmt. 16:109-126. Bun-ell. V. G. 1985. Oyster culture, pp. 235-273. In: J. V. Huner and E. E. Brown (ed.). Crustacean and Mollusc Aquaculture in U.S.A. Avi Publishing, Connecticut. Bustillos. J. & E. Olivares. 1986. Fitoplancton en tres ecosistemas de manglar de Bahia de La Paz, B.C.S. (Mexico). II Produccion pnmaria bruta. neta y respiracion. .Anal. Inst. Ciencias Mar Limnol. UNAM 13:301-306. Caceres. C, D. Ramirez, G Chavez & O. Peiialoza. 1986. Diseiio y manejo de costales y estanque para el cultivo de ostiones. (unpub- lished) U.A.B.C.S., Mexico. Caddy. J. F. 1989 Overview of asessment and management of marine molluscs and invertebrates other than crustaceans, pp. 397^00. In: Production. Growth, and Survival of C. Gigas 607 J. F. Caddy (ed). Manne Invertebrate Fisheries: Their Assessment and Management. John Wiley. New York. Chapman. D. W. 1978. Production, pp. 202-217. In: T. Bagenal (ed.). Methods for Assessment of Fish Production in Fresh Waters. IBP. Handbook No. ?>. Blackwell. Oxford. Cnsp. D. J. 1984. Energy flow measurements, pp. 197-279. In: N. A. Holme and A. D. Mclntyre (ed.l. Methods for the Study of Marine Benthos. IBP. Handbook No. 16. Blackwell. Oxford. Fidalgo. J., A. Cid. Y. Lopez-Munoz. J, Abalde & C, Herrero 1994. Growth and biochemical profile of juvenile mussels (Myliliis gallopm- vincialis Lmk) fed on different algal diets. J. Shellfish Res. 13:67-75. Flores. S. R. 1994. Utilizacion ecologica de Bahia de La Paz. B.C.S. por el rorcual tropical Balaenoplera edeni. 1988-1991. Tesis de Maeslria en Ciencias. U.A.B.C.. Mexico. Fujiya. M. 1970. Granjas pesqueras en Japon. Memoradum tecnico No. 6-02-1. S.R.H.. Mexico, pp. I-I29. Gardner. J. P. & M. Thomas. 1987. Growth, mortality and production of organic matter by a rocky intertidal population of Mylinis ediilis in the Quoddy Region of the Bay of Fundy. Mur. Ecol. Progr. Ser. V 39: 31-36. Golikov. A. N. & V. Menshutkm. 1973. Estimation of production prop- erties of mollusk populations. Mar. Biol. 20:6-13. Hall. D J. 1971. The experimental field approach to secondary produc- tion. In: W. T, Edmonson and G. G. Winberg (ed.). A Manual on Methods for the Assessment of Secondary Productivity in Fresh wa- ters. Blackwell Scientific Publishers, Oxford Hall, S. 1984. A multiple regression model for oyster growth. Fish. Res. 2:167-175. Islas. R. O. 1975. El ostion japones Crassosirea gigas en Baja California. Ciencias Marinas 2:50-59. Islas, R. O., V. Guajardo & A. Perez. 1992. Crecimiento y sobrevivencia del ostion japones (Crassostra gigas) en la Laguna Manuela: B.C., Mexico. Ciencias Marinas 8:31-38. Kimmerer. W. J. 1987. The theory of secondary production calculations for continuously reproducing populations. Limnol. Oceanogr. 32:1- 13. Kinne, O. 1978. Manne Ecology. 4 vols. John Wiley, New York. Korringa, P. 1976. Farming the Cupped Oysters ef the Genus Crassosirea. vol. 2. Elsevier, Amsterdam. Langdon, C. & R. Newell. 1990. Utilization of detritus and bacteria as food sources by two bivalve suspension-feeders the oyster Crassoirea virginica and the mussel Geukensia demissa. Mar. Ecol. Progr. Ser. V 58:299-310. Lango, F. 1994. Estudios basicos sobre depredadores activos y poten- cialespara el desarrollo del cultivo extensivo de Argopecten circularis. Tesis de Maestri en Ciencias. C.I. C.I. MAR.. Mexico. Lechuga. C. 1977. La produccion organica primaria en Ensenada de La Paz. Centro de Investigaciones Biologicas, La Paz. Mexico, pp. 47- 69. Mathews. C. P. 1970. Estimates of production with reference to general surveys. OWOS 2 1 : 1 29- 1 33 . Morin. A.. M. Constantin & R. Peters. 1988. Allometric models of sim- uliid growth rates and their use for estimation of production. Can. J . Fish. Aqiiat. Sci. 45:315-324. Morin, A.. T. A. Mousseau & D. A. Roff. 1987. Accuracy and precision of secondary production estimates. Limnol. Oceanogr. 32:1342-1352. Muhlia. A.. R. Olson & W. Hazen. 1980. Analisis comparative de los parametros de crecimiento de tres especies de bivalvos del genero Macoma. en el Paci'fico Nororiental. IV Congreso Nacional de Ocean- ografia. Mexico. Munro. J. L. 1989. Fisheries for giant clams (tridacnidae: bivalvia) and prospects for stock enhancement, pp. 541-558. In: J. F. Caddy, (ed.) Marine Invertebrate Fisheries: Their Assessment and Management. John Wiley, New York. Negus. C, L. 1966. A quantitative study of a growth and production of unionid mussels in the river Thames at Reading. J . Aniin. Ecol. 35: 513-532. Paez. F.. H Zazueta & J. Osuna. 1993. Biochemical composition of the oysters Crassosirea iridescens Hanley and Crassosirea corleziensis Hertlein in the Northwest coast of Mexico: seasonal changes. J Exp. Mar. Biol. Ecol. 170:1-9. Parsons. T. R.. M. Takahashi & B. Hardgrave. 1984. Biol, Oceanogr. Processes. Pergamon Press. Oxford. Reinitz. G. & T. Yu. 1981. Effects of dietary lipids on growth and fatty acids composition of rainbow trout [Salmo gairneri). Aquaculntre 22: 359-366. Rcnard. P 1991 Cooling and freezing tolerances in embryos of the Pacific oyster. Crassosirea gigas: methanol and sucrose effects. Aquacullure 92:43-57. Richer. W. E. 1975. Computation and interpretation of biological statistics offish populations. Bull. Fish. Res. B Can. 191. Rigler. F. & J. Downing. 1984. The calculation of secondary productivity. In: J. Downing and F. Rigler (ed.). A Manual of the Methods for the Assessment of Secondary Production in Fresh Waters. 2nd ed IBP Handbook. No. 17. Blackwell. Oxford. Robert. R.. G. Trut. M. Borel & D. Maurer. 1993 Growth, fatness and gross biochemical composition of the Japanese oyster. Crassosirea gigas. in Stanway cylinders in the Bay of Arcachon. France. Aqua- cullure 110:249-261. Ruiz.C. &C. Caceres. 1990. Estudio preliminarde captacion de juveniles de moluscos bivalvos en Bahi'a de La Paz. Mexico. Invesl. Marinus CICIMAR 5:29-38. Santoyo, H. 1994. Fitoplancton y Productividad. pp. 221-245. In: G, De la Lanza and C. Caceresa (ed). Lagunas Costeras y el Literal Mexi- cano, UABCS. BCS, Mexico. Shafee, M. & M. Sabatie. 1986. Croissance et mortalite des huitres dans la lagune de Oualidia (Maroc). Aquacullure 53:201-214. Signoret, M. & H. Santoyo. 1980. Aspectos ecologicos del plancton de Bahia de La Paz. Anal. C. Ciencias Mar Limnol. UN AM 7:217-248. Statgraphics. 1986. Statistical Graphics System. Statistical Graphical Cor- poration, Rockville, MD. Sukhotin. A. & N. Maximovich. 1994. Variability of growth rate in Mylitus edulis L. from the Chupa Inlet (the White Seal. J. E.xp. Mar. Biol. Ecol. 176:15-26. Thompson. R. J. 1984. Production, reproductive effort, reproductive value and reproductive cost in a population of the blue mussel Mytilus edulis from subarctic environment. Mar. Ecol. Prog. Ser. V 16:249- 257. Valiela. I. 1984. Marine Ecological Processes. Springer Vedag, New York. Villamar, A. 1965. La fauna malacologica de la Ensenada de La Paz con notas ecologicas. An. Insl. Nat. Inv. Biol. Pesq. Me.xico 1:115-152. Viola. S.. S. Mokady, D. Behar & U. Cogan. 1988. Effects of polyun- saturated fatty acids in feeds of tilapia and carp 1 . Body composition and fatty acid profiles at different environmental temperatures. Aqua- cullure 15:m-]31. Walne, P. R. & B. E. Spencer. I97I. The introduction of the Pacific Oyster (Cras.iostrea gigas) into the United Kingdom Shellfish Infor- mation Leanet, MAFF. Fisheries Experiment Station Conway, North Wales, United Kingdom. March 1971. Warwick, R. M. 1980. Population dynamics and secondary production in benthos, pp. 1-24. In: R. Tenore and C. Coull (ed). Manne Benthic Dynamics. University of South Carolina Press, Columbia. Zaika, V. E. 1973. Specific Production of Aquatic Invertebrates. John Wiley and Sons, New York. Zar, J, H. 1996. Biostatistical Analysis. Prentice Hall, Englewood Cliffs, NJ. Journal of Shellfish Research. Vol. 15. No. 3. 609-615, 1996. STUDIES ON TRIPLOID OYSTERS IN AUSTRALIA. VII. ASSESSMENT OF TWO METHODS FOR DETERMINING TRIPLOIDY IN OYSTERS: ADDUCTOR MUSCLE DIAMETER AND NUCLEAR SIZE CALEB GARDNER,* ' GREG B. MAGUIRE, AND GREG N. KENT Department of Aqiiaciilture University of Tasmania Launceston, Tasmania 7250. Australia ABSTRACT Two methods for distinguishing tnploid Pacific oysters [Crassoslrea gigas (ThunbergI] from diploid oysters were assessed. Adductor muscle diameter in relation to valve height was significantly (p < 0.001 ) greater in samples of tnploid oysters than in diploid samples and was influenced by site. However, variation in this measure was too large to allow individual oysters to be distinguished as either tnploid or diploid. A second method was assessed that used differences in the nuclear size of hemocytes and the intensity of staining of hemocyte nuclei to distinguish between diploids and tnploids. Histological sections, prepared by standard paraffin histology, were stained for nuclear histones with Gill's hematoxylin. Integrated nuclear optical density and nuclear area were recorded with image analysis. This method was effective in distinguishing individual oysters as diploid or tnploid. When histological specimens are required, this method is less expensive than other techniques used to determine tnploidy. KEY WORDS: Triploid, oysters, adductor muscle, nuclei, image analysis INTRODUCTION In any attempt to produce triploid bivalves, it is necessary to test to what degree this has been achieved, because generally, some diploid individuals are also produced. The ratio of diploids to tnploids is established to save wasted labor rearing batches of larvae when the proponion of triploids is low (Beaumont and Fairbrother 1991), spat before sale, broodstock for the production of tetraploids (Guo and Allen. 1994), and research of triploid peiformance. Where histologic data are required from these oys- ters, it would be useful to also establish ploidy by histology. In an evaluation of the performance of triploid Pacific oysters, Crassostrea gigas (Thunberg), at commercial leases in Tasmania (Gardner et al. in prep.. Maguire et al. in prep.), it was necessary to establish the ploidy of adult oysters sampled for the histology of gonad development. In concurrent samples, the proportion of dip- loid oysters in the triploid group was assessed by the use of flow cytometry, although these samples did not include oysters pro- cessed for histology. It was found that approximately 257c of the "triploid" group were diploid oysters. Individuals among the pu- tative triploids that developed large gonads may have been dip- loids; alternatively, they may have been triploids where gameto- genesis was less inhibited. Therefore, we attempted to develop another indicator of triploidy that could be used to retrospectively assess samples already processed for histology. Histologic sam- ples could then be used to accurately describe the gametogenesis of triploid oysters and to compare sexes in relation to the suppres- sion of gametogenesis (Gardner et al. in prep.). Several techniques have been described for the determination of the ploidy of bivalves; the two approaches most frequently used are karyotypic analysis and flow cytometry. Karyotypic analysis involves the preparation, staining, and counting of the chromo- somes of separate nuclei (Kilgerman and Bloom 1977, Allen 1983). Although karyotypic analysis is an accurate measure of ploidy, it is very time consuming (Komaru et al. 1988). Flow *To whom correspondence should be addressed. Present address: Division of Manne Resources. Taroona Research Lab- oratones, GPO Box 192B Hoban, Tasmania 7001, Australia. cytometry is capable of recording the DNA content of cells at a much greater rate, in the order of 10,000-100,000 nuclei per sam- ple. This technique is particularly useful for the estimation of the percentage of triploids in a population, as is required in hatcheries (Chaiton and Allen 1985). The purpose of this study was to retrospectively verity the ploidy of oysters that had been processed for histology, and also to evaluate techniques that could be used for the assessment of indi- vidual oysters in subsequent studies. This was approached in two ways. First, a morphological indicator, adductor muscle diameter in relation to whole body size, was used; it was anticipated that this technique, if effective, could be used on farms. Second, the size and staining intensity of nuclei in histologic sections were assessed. The first approach, measurement of body structure, was as- sessed by measuring the size of the adductor muscle in relation to the size of the whole organism. It was hypothesized that the in- creased DNA content of muscle fiber nuclei may be reflected in the diameter of muscle cells. This in turn may affect the diameter of the entire adductor muscle and an increase in adductor muscle diameter. To some extent, the large nuclei may increase the size of the whole organism (polyploid giantism), so that there would be no increase in the relative size of the adductor muscle. However, an increase in adductor muscle size from triploidy has been re- ported for Sydney rock oysters Saccoslrea commercialis (Iredale and Roughley) (Nell et al. 1994). Factors other than nuclear size may influence the observed increase in adductor muscle size, such as increased storage of nutrients in the adductor muscle as a result of reduced gametogenesis (S. K. Allen, Jr., Rutgers University, pers. comm.). The additional set of chromosomes within the nucleus of trip- loid cells can be expected to alter the size of the nucleus and/or the density of the DNA contained therein. A simple increase in nu- clear size is not always sufficient to distinguish ploidy (Jarvis 1992a); however. Child and Watkins (1994) were able to distin- guish triploid Manila clams \Tapes philippinarum (Adams and Reeve)] by the larger diameter of the nucleus of gill tissue cells in comparison to diploids. The amount of DNA in the nucleus can be measured by staining the DNA with a specific fluorescent dye, and 609 610 Gardner et al. the degree of fluorescence is then quantified by flow cytometry (Coon and Weinstein 1992). Image analysis uses a similar tech- nique to quantify the DNA content of nuclei. Instead of measuring the light emitted by stained nuclei (as in flow cytometry), image analysis can be used to measure the light absorbed by the stained nuclei (Jarvis 1992a). By measuring the light absorbed by nuclei stained by the Feulgen method as a function of their area (as integrated optical density). Gerard et al. (1994) independently used image analysis to distinguish diploid oysters from triploids. Nuclear size and integrated optical density were calculated for commercial-size oysters (approximately 60 g) from experimental groups of triploid Pacific oysters sampled in the months before and after a spawning recorded at Birch's Bay (southern Tasmania. Australia), December 1992 and January 1993. and comparisons were made with triploid groups at two other sites. This allowed an evaluation of individuals within the triploid group that had gonad morphologies similar to those of diploids. Were these individuals triploids in which gametogenesis was not inhibited, or were they simply diploids? MATERIALS AND METHODS Sampling Sites The diploid and triploid Pacific oysters used in these experi- ments were from the same groups as those described by Maguire et al. (199-) and were grown at three sites in Tasmania: Little Swanport (east; 148°00'. 42°19'). Pitwater (southeast: 147°3()'. 42°50'). and Birch's Bay (south: 147°I5'. 43°10'). These groups arose from the same pool of gametes. Triploids were induced by the suppression of polar body 2 with cytochalasin B (Allen et al. 1989) in February 1990. and spat were grown to market size intertidally. Adductor Muscle Adductor muscle size index was calculated by dividing adduc- tor diameter (in millimeters) by valve height (in millimeters), with measurements taken from 81 diploid and 86 triploid oysters. The measurement of adductor muscle diameter was taken from the upper valve along the axis of the border between quick and catch (opaque and translucent) muscle. Oysters for adductor muscle measurements were cultured in a replicated trial with three repli- cates of 10 oysters per ploidy group at Birch's Bay and four replicates at each of the other sites (10 oysters per replicate at Pittwater. 3 per replicate at Little Swanport). Sampling occurred at approximately 90-mm valve height, which was reached in 29 mo at Little Swanport and Pittwater and in 42 mo at Birch's Bay. The effects of ploidy. site, and interaction of ploidy with site on ad- ductor muscle size index were assessed with a two-way fixed- factor analysis of variance after an assessment of the homogeneity of variance (Cochran's test) and normality (Shapiro-Wilk test) (Walpole and Myers 1989). Nuclear Measurements Specimens for nuclear measurements were collected from Birch's Bay in December 1992, at the peak of gonad development, and also in January 1993, after spawning had occurred in diploid groups. Standard paraffin sections (5 (jim) were prepared from a slice of tissue taken slightly above the junction between the labial palps and the gills (Morales- Alamo and Mann 1984). The propor- tion of somatic tissue (excluding gills) occupied by gonad, and expressed as percent gonad area, was determined by the use of image analysis from histologic sections, as described by Maguire etaL (199-). Standard paraffin (7-fjim) histologic sections of oysters for im- age analysis were prepared as described in Gardner et al. (in prep. I. The staining of tissue for image analysis must be specific and of high quality (Jarvis 1992b). Jarvis ( 1992b) suggests that the Feulgen technique is the most appropriate stain for the image analysis of DNA. This was not found to be the case in this study, because Feulgen staining produced a pale stain that provided poor contrast against the surrounding tissue. Variation in staining in- tensity between slides was also encountered with the Feulgen tech- nique, and this masked variation due to differences of ploidy. The source of variation in Feulgen staining was considered to arise from a critical step of acid hydrolysis. Consequently, an alterna- tive stain was sought that produced strong and specific staining of DNA and that also was very simple so that variation was dimin- ished. Gill's hematoxylin (Gill et al. 1974) was found to be a suitable stain. Sections were stained for precisely 2 min and then were rinsed with tap water for 5 min. Although Feulgen staining was inconsistent in this study, Gerard et al. (1994) used Feulgen staining to accurately distinguish triploids by image analysis. Nuclear area and densiometric (integrated optical density) mea- surements were taken with a BH2. Olympus'^ compound micro- scope. Slides were viewed with a lOOx planachromat objective with oil immersion to produce a magnification of x 1000. Nuclear area and integrated optical density measurements were recorded for 20 hemocytc cells from each specimen. Measurements were taken with a CUE 11®. lBM®-compatible. image analysis system. Heniocyle cells were chosen because they were widely distributed through the tissue and there appeared to be little variation of hemo- cyte nuclear size within specimens. Also, the nuclei of hemocytes are round, which assists in defining the shape to be measured with image analysis. Child and Watkins ( 1994) also used hemocytes to facilitate the measurement of cell area. Illumination between sections was standardized by adjusting the light intensity of the microscope to a standard intensity. This was done when viewing a section of each slide without tissue. The purpose of this step was to remove variation in light intensity due to differences in the slide or cover slip thickness and variation in the optical density of the mounting medium. Densiometric mea- surements were made with microscope adjustment and illumina- tion filtration and illumination adjustment, as advised by Jarvis 11992b). Basophilic components within the cytoplasm produced pale staining, which compounded nuclear optical density measure- ments. To correct for this, an optical density reading (from an area the exact size of the nucleus ) was taken from the cytoplasm of each hemocytc sampled. By subtracting the integrated optical density of the cytoplasm from the integrated optical density of the nucleus, this component of light absorption was removed. The significance of differences in nuclear measurements between triploids and dip- loids was tested with a Wilcoxon nonparametric test (Walpole and Myers 1989). RESULTS AND DISCUSSION Adductor Muscle The adductor muscle diameter in relation to valve height (ad- ductor muscle index) was significantly greater (p < 0.001) in triploid oysters than in diploids at all sites (Table I). This index was 9.9% greater in triploids, and Nell et al. (1994) found that the Determination of Triploidy in Oysters 611 TABLE 1. Mean adductor muscle size index (diameter expressed as a percentage of valve height ± SE) for diploid and triploid Pacific oysters (C. gigas) from three sites in Tasmania, .Australia. Site Diploid Group Triploid Group Significance of Effect of Ploidy Little Swanporf" I'lttwater'' Birch's Bay" 20.9 (±0.49) 17.2 (±0.17) 17.4 (±0.69) 22.7 (±0.77) 19.2 (±0.29) 19.2 (±0.35) <0.001 <0.00I <0.001 Note: Replication at Little Swanport and Pittwater = lour replicates of 10 individuals and at Birch's Bay = three replicates of 10 individuals. Sig- nilicant dittcrcnce between sites (diploids and tnploids combined) for ad- ductor muscle index ib denoted by a difference in superscnpt (p < 0.05). The tnploid samples were from populations that contained tnploid and diploid oysters in about a 3:1 ratio. equivalent value for Sydney rock oysters held subtidally was 5.7%. Although there was a significant difference in adductor muscle index between diploids and triploids. there was extensive overlap of values between ploidy groups at all three sites (Fig. 1 ). Consequently, it was considered impossible to distinguish diploids from tnploids by the use of this index. The measure of adductor muscle size that was used in this study was adductor muscle diameter, which was measured with callipers and could have been used on farms. More sensitive measures, such as adductor muscle area or overall weight, would require image analysis or precision balances and are unsuitable for farms. Al- though unsuitable for the objectives of this study, these more sensitive measures may be effective and useful for other situations. Although ineffective for determining the ploidy of individual oysters, the larger ratio of adductor muscle diameter to valve height in triploids may be beneficial in aquaculture to enhance survival through periods of prolonged exposure. However, this index would be sensitive to any differences in shell shape between diploids and triploids. Site was also found to influence adductor muscle index (p < 0.001), with the largest values recorded from Little Swanport. Oyster racks at this site were exposed for much longer periods than at the other two sites (Maguire et al. in prep.). The interaction between site and ploidy was not significant (p > 0.05). 4- &^ 3- 3 S- 2. 1- a. Little Swanport I E] Diploid ■ Triploid i U -j — I — I — I — I — I — I — I I I I I I I I r 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Adductor Diameter (% of shell height) T — I — r 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Adductor Diameter (% of shell height) T 28 T I I I I I ~ 15 16 17 18 19 20 21 22 23 24 25 26 Adductor Diameter (% of shell height) Figure 1. Frequency of adductor muscle index of triploid and diploid Pacific oysters, C. gigas. cultured at three sites in Tasmania, Austra- lia. (Triploid samples were from populations that contained triploid and diploid oysters in about a 3:1 ratio,) Image Analysis A highly significant difference (p < 0.001 ) was found between diploids and triploids for both nuclear diameter and integrated optical density, as assessed by image analysis (Table 2). This was the case for both December and January samples at Birch's Bay. This demonstrates the potential for hemocyte nuclear area and TABLE 2. Comparison of nuclear area and integrated optical density of hemocyte nuclei between triploid and diploid Pacific oysters for December 1992 and January 1993, Integrated Optical Density: Mean ± SO (n) Nuclear Area ((im Mean±SD(n) '»: Ploidy December January December January Diploid Triploid 0.87 ± 0.06(22)-" 1.09 ± 0.17 (20)" 0.80 ± 0.09 (21)- 1.03 ± 0.15 (21)" 5.25 ± 0.40 (22)-' 6.22 ± 0.58 (20)" 4.46 ± 0.25 (21)" 5.73 ± 0.61 (21)" Note: Triploid group contains tnploid and diploid oysters in about a 3:1 ratio. Data are intended to convey differences due to ploidy within monthly samples. Comparison of means between monthly samples may be compounded by error from differences in the set-up of the imaging system between the two samples. Significance of difference between ploidy groups is denoted by a difference in superscript (p < 0.001). 612 Gardner et al. 80 60- T3 § 40 O 20- (a) 7 9 nV ■7.5 - 7 6 S CM 6 =1 6 c« aj Lh < 5.5 D O 3 5 z -4.5 Hn : 13 15 17 19 21 23 25 27 29 31 33 35 37 39 Specimen -1.45 o (N +1 - 1.2 ^ C Q C3 -0.95 ex O T3 15 17 19 21 23 25 27 29 31 33 35 37 39 0.7 Specimen Figure 2. Mean hemocvte nuclear area (a) and integrated optical density (bl in relation to the proportion of somatic tissue occupied bv gonad for the December 1992 sample. Gonad area values are represented by columns, and nuclear parameters are represented by solid triangles. Specimens 1-20 are from the diploid group; specimens 21^0 are from the triploid group. No standard deviations were recorded for nuclear area measurements from this sample. integrated optical density to be effective tools in the determination of ploidy; however, for the purpose of this study, it was necessary to distinguish diploid and triploid oysters on an individual basis rather than by simply separating populations. The degree to which this was achieved was assessed by comparing values obtained by image analysis against the gonad area of individual oysters (Figs. 2 and 3). Both nuclear area and integrated optical density values ap- Determination of Triploidy in Oysters 613 CO c o a 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 Specimen -a CO C o O -1.45 - 1.2 -0.95 o T3 C/3 +1 Figure 3. Mean for the January Specimens 1-20 1/3 G Q o G- O T3 (U C3 l-i 00 (D C 0.7 ^ Specimen hemocyte nuclear area (a) and integrated optical density (bl in relation to the proportion of somatic tissue occupied by gonad 1993 sample. Gonad area values are represented by columns, and nuclear parameters are represented by solid triangles, are from the diploid group; specimens 21^1 are from the triploid group. peared to broadly distinguish triploid oysters from diploid oysters in the December 1992 sample. Neither measurement conclusively separated all individuals, but a clear indication of ploidy was ap- parent when both measures were compared for each oyster. Some triploid oysters had nuclear area measurements that were higher than those of most diploids but that were too low to clearly indicate them as triploid (e.g., specimens 23 and 24, Fig. 3). In these cases, the integrated optical density measurements clearly inai- cated the oysters to be triploid. Morphometric measurement of nuclear size alone was not sufficient to separate all diplo; and triploid Pacific oysters, although this is possible with Alanila clams (Child and Watkins 1994). The relationship of each of the 614 Gardner et al. TABLE 3. Comparison of the variation (by coefficient of variation) between nuclear area and integrated optical density measures for diploid and triploid oysters. Integrated Optical Density: Coefficient of Variation (n) Nuclear Area: Coefficient of Variation (n) Ploidy December January December January Diploid 797(22) 17.01(211 7.20(22) b. 32(21) Tnploid 8.69(20) 11.64(21) 9.96(20) 8.40(21) Note: Oysters within the triploid group that appeared to be diploids were excluded from analyses. Coefficient of variation = (standard deviation/ mean) x 100. nuclear parameters (integrated optical density and nuclear area) to ploidy was based on both the gonad area measurements of each oyster and the ratio of diploids within the tnploid group, as de- termined by flow cytometry. The nuclear area and integrated op- tical density values indicated that eight oysters within the triploid group, from both months (a total of 40 triploids), were diploid, which is equal to 20% of the sample (specimens 22. 28. and 32 from the December sample, specimens 21. 32. 36. 37. and 3X from the January sample). Although the sample size is very small, this is in the order of what would be expected on the basis of the original ratio obtained by flow cytometry (approximately 25% diploids within triploid samples). There tended to be more variation in integrated optical density measurements than in nuclear area data (Table 3). The nuclear staining capacity of hematoxylins has been attributed to the bind- ing of the dye molecule to nuclear histones (Stevens 1982). but other proteins will also bind these dyes. The variation in integrated optical density measurements observed in this study may have been caused by the nonspecific staining of proteins associated with the nucleus, rather than the staining of histones alone. Staining intensity can be expected to vary from day to day, and there appeared to be some difference in intensity between the December and January samples, which were stained separately (Figs. 2b and 3b). To eliminate error caused by changes in the staining of the hematoxylin, diploid controls are necessary for each batch of anal- yses. Although uniform and intense Feulgen staining could not be achieved in this study, it has been used elsewhere for the deter- mination of ploidy (Jarvis et al. 1992b. Gerard et al. 1994) and it is highly specific to DNA. Variations on the Feulgen technique, used for determining integrated optical density in human pathol- ogy (Schieck et al. 1987. Schulte et al. 1988). may improve the staining intensity in oyster sections and further assist in determin- ing ploidy. The nuclear area and integrated optical density values of trip- loids from the December 1992 sample confirmed the suppression of gametogenesis as a result of triploidy. Those individuals that did develop extensive gonad were shown to be diploids. In the following sample. January 1993, there were three individuals within the triploid group with large gonad area that appeared to be legitimate triploids on the basis of image analysis (specimens 22. 24. and 33. Fig. 3). All of these triploids were male, and gonad size is less retarded in male triploid Pacific oysters than in female triploids (Allen and Downing 1990). Numerous spermatocytes were present in the follicles of these individuals, yet no sperma- tozoa were observed, which is a pattern of gametogenesis consis- tent with that reported for triploid Crassoslrea vir^inica (Gmelin) (Barber and Mann 1991). The technique of image analysis of hemocyte nuclei allowed us to retrospectively determine ploidy and to separate these groups for an analysis of gametogenesis. Where histologic sections are also required, the technique of image analysis provides a useful means of determining ploidy and may also serve as an adjunct to flow cytometry. ACKNOWLEDGMENTS Financial assistance from the University of Tasmania and the Fisheries Research and Development Corporation is gratefully ac- knowledged. Thanks are also due to Dr. B. Munday, Dr. S. Allen, and Dr. J. Nell for their comments on the manuscript and to Dr. P. Baird for assistance during the inception of the image analysis component. LITERATURE CITED Allen, S. K. Jr. 1983. Flow cytometry: assaying expenmental polyploid fish and shellfish. Aqiuuulture 33:317-328. Allen, S. K. Jr. & S. L. Downing. 1990. Performance of tnploid Pacific oysters, Crassoslrea gigas: gametogenesis. Can. J. Fish. Aijuat. Sci. 47:1213-1222. Allen, S. K. Jr.. S. L. Downing & K. K. Chew. 1989. Hatchery Manual for Producing Tnploid Oysters. University of Washington Press, WA. 27 pp. Barber. B. J. & R. Mann. 1991. Sterile triploid Crassoslrea virginica (Gmelin, 1791 ) grow faster than diploids but are equally susceptible to Perkinsus marinus. J. Shellfish Res. 10:445^50. Beaumont, A. R. & J. E. Fairbrother. 1991. Ploidy manipulation in mol- luscan shellfish: a review. J. Shellfish Res. 10:1-18. Chaiton, J. A. & S. K. Allen, Jr. 1985. Early detection of triploidy in the larvae of Pacific oysters, Crassoslrea gigas by flow cytometry. Aqua- culture 48:35^3. Child, A R. & H. P. Watkins. 1994 A simple method to identify triploid molluscan bivalves by the measurement of cell nucleus diameter. Aquaculture 125:199-204. Coon, J. S, & R. S. Weinstein. 1992. Specimen preparation, cell mea- surements and probes for flow cytometry, pp. 27-41. In: Proceedings of the inaugural Scientific Meeting; Introduction to Quantitative Diag- nostic Pathology. Australian Association for Quantitative Pathology. Melbourne. Victoria. Gardner. N. C, M. S. R. Smith, G. B. Maguire, G. N. Kent & J. A. Nell. Studies on tnploid oysters in Australia. III. Gametogenesis of tnploid and diploid Pacific oysters, Crassoslrea gigas (Thunberg) in Tasmania. (Submitted \.o Aquacullure). Gerard, A., Y. Nacin, J. M. Peignon, C. Ledu. P. Phelipot, C. Nouet, I. Peudenier & H. Grizel. 1994. Image analysis: a new method for esti- mating triploidy in commercial bivalves. Aquacull. Fish. Manage. 25:697-708. Gill, G. W.. J. K. Frost & K. A. Miller. 1974 A new formula for a half oxidized hematoxylin solution that neither overstains nor requires dif- ferentiation. Ada Cxtot. 18(4):3OO-310. Guo, X. & S. K. Allen. Jr. 1994. Viable tetraploids in the Pacific oyster i.Crassoslrea gigas Thunberg) produced by inhibiting polar body 1 in eggs from triploids. Mol. Mar. Biol. Biotechnol. 3:42-50. Jarvis. L. R. 1992a. The microcomputer and image analysis in diagnostic pathology. Microsc. Res. Tech. 21:292-299. Jarvis. L. R. 1992b. Practical aspects of video image analysis for densi- Determination of Triploidy in Oysters 615 ometric measurement, pp. 99-115. In: Proceedings of the Inaugural Scientific Meeting; Introduction to Quantitative Diagnostic Pathology. Australian Association for Quantitative Pathology. Mclhoume. Victo- ria. Kilgerman. A. D. & S. E. Bloom. 1977. Rapid chromosome preparations from solid tissues of fishes. J. Fish. Res. Bd. Can. 34:266-269. Komaru. A., Y. Uchimura. H. Leyama & K. T. Wada. 1988. Detection of induced triploid scallops. Chlamys nobilis. by DNA microtluorometry with DAPl staining. .Aquacuhure 69:201-209. Maguire. G. B., N. C. Gardner. J. A. Nell. G. Kent & A. Kent. Studies on triploid oysters in .Australia. II. Growth, condition index, glycogen content and gonad area of triploid and diploid Pacific oysters, Cras- sosirea aigas (Thunbergl. in Tasmania. (Submitted to Aqiiuciilitire). Morales-Alamo. R. & R. Mann. 1989. Anatomical features in histological sections of Cnis.wslrea virginim (Gmelin. 1791 ) as an aid in measure- ment of gonad area for reproductive assessment. J. Shellfish Res. 8: 71-82. Nell. J. A.. E. Co.\. I. R. Smith & G. B. Maguire. 1994. Studies on tnploid oysters in Australia. 1. The farming potential of triploid Sydney rock oysters Saccoslreu commercialis (Iredale and Roughley). Aqua- culture 126:243-255. Schieck. R.. G Taubert & H. Krug. 1987. Dry mass. DNA and non- histone protein determinations in lung cancer cells. Hisiochem. J. 19: 504-508. Schulte. E.. D. Wittekind & V. Kretschmer. 1988. Victoria blue B— a nuclear stain for cytology. Hislochemislry 88:427^33. Stevens. A. 1982. The haematoxylins. pp. 109-121. In: J. D. Bancroft and A. Stevens (eds.). Theory and Practice of Histological Techniques. 2nd ed. Churchill Livingstone. London. Walpole. R. E. & R. H. Myers. 1989. Probability and Statistics for En- gineers and Scientists, 4th ed. Macmillan. New York. Journal nf Shellfish Research. Vol. 15. No. 3, 617-622, 1996. ANNUAL PATTERN OF BROODING IN POPULATIONS OF CHILEAN OYSTERS, TIOSTREA CHILENSIS, (PHILIPPI, 1845) FROM NORTHERN NEW ZEALAND A. G. JEFFS,'"^ R. G. CREESE,^ AND S. H. HOOKER^ ^Cawthron Institute Private Bag 2 Nelson, New Zealand 'Leigh Marine Laboratory University' of Auckland Private Bag 92019 Auckland. New Zealand School of Environmental and Marine Sciences University of Auckland Private Bag 92019 Auckland, New Zealand ABSTRACT The annual pattern of brooding in two populations of adult Chilean oysters, Tiostrea chilensis. in northern New Zealand, was examined. Both populations were brooding larvae throughout the year, with less brooding activity in winter and increased larval production around spring and early summer. Despite the extended brooding season, the proportion of individuals brooding during peak periods remained high. Larger broods of larvae appeared to be associated with periods of raised brooding activity and lower water temperatures. In both populations, there were no differences in the size of oysters brooding at different times of the year. Overall, the annual pattern of brooding in both populations was markedly different from previous reports for this species in other regions. These differences tend to confimi the role of water temperature as the most important environmental factor regulating the annual pattern of reproduction in this oyster species. This conclusion has important implications for the hatchery production of larvae for aquaculture. KEY WORDS: Chilean oyster, Tiosirea chilensis. reproductive cycle, brooding. New Zealand, flat oyster, Ostreidae INTRODUCTION The Chilean oyster Tiosirea chilensis {PhiVippi, 1845) is a com- mercially important flat oyster that is native to New Zealand and the Pacific coast of South America (Osorio 1979, Beu and Max- well 1990. Jeffs and Creese 1996). This oyster, along with all species of the genus Ostrea. incubates its eggs after fertilization and releases larvae (Roughley 1929, Millar and Mollis 1963). Like some other species of flat oyster, Tiostrea is a protandrous her- maphrodite, maturing first as a male, and later in life, as a female (Mollis 1963. Winter el al. 1984. Chanley and Chanley 1991). However, unlike all other species of oyster, it can brood its larvae through their entire development (Millar and Mollis 1963). Studies of the annual breeding cycle in Tiostrea, have generally found a small proportion of adult oysters brooding larvae for only a short period during the spnng and/or summer (Tunbridge 1962. Stead 1971. Cranfield and Allen 1977. Westerskov 1980. Winter et al. 1984). Some workers have hypothesized that breeding in Tiostrea is regulated by water temperature, and therefore, populations at lower latitudes could be expected to have more extensive brooding seasons (Westerskov 1980. Cranfield and Michael 1989). The aim of this study, therefore, was to investigate the annual pattern of brooding in populations of Chilean oysters from lower latitudes in New Zealand and to compare the results with previous research conducted at higher latitudes, in both New Zealand and Chile. MATERIALS AND METHODS Two populations of Chilean oysters in the north of the North Island of New Zealand were used for this study; Te Tau Bank in the Manukau Harbour (37°02'S.I74°4rE), a shallow harbour on the northwestern coast, and Moturekareka Island in the Hauraki Gulf (36°28.50'S,174°47.60'E). a large embayment on the north- eastern coast (Fig. 1). Around 75 oysters, of a size known to be capable of brooding larvae (Jeffs, unpublished data), were sam- pled haphazardly from each population at approximately monthly intervals. Sampling began in December 1992 from the Manukau Harbour and in April 1994 from the Hauraki Gulf. For both pop- ulations, sampling continued until December 1995. The shell height of the oysters in each monthly sample was measured to the nearest millimeter with vernier callipers before they were opened. Larvae in brooding oysters were removed with a wash bottle and then counted in a manner similar to that described by Walne (1964). An assessment of this method confirmed that it provided an unbiased estimate of the size of broods with a higher degree of accuracy than has been reported previously. A mercury thermom- eter was used to record water temperatures to the nearest 0.1 °C during a series of visits each month to both study sites. Results Annual Pattern of Brooding — Manukau Harbour Thirty-six collections of oysters were made between December 1992 and December 1995. In total. 2.635 adult oysters were ex- amined, of which 176 (6.7%) were found to be brooding. Brood- ing oysters were found in every month of sampling, with the exception of January 1993. showing that breeding takes place year-round (Chart in Fig. 2). The highest proportions of brooders 617 618 Jeffs et al. Figure 1. Map showing the location of the two study populations of T. chilensis in northern New Zealand and other locations mentioned in this article. in all of the monthly samples were in September 1993 ( 18.2%) and in October 1995 (16.9%). In each of the 3 y. lower proportions of oysters tended to brood around January and July and higher pro- portions tended to brood between September and November. Some samples from within the period of March to May of each year also showed elevated proportions of brooding oysters that perhaps corresponded with a secondary breeding season. The an- nual pattern of larval production for the population was estimated by dividing the total number of larvae found in each monthly sample by the total number of oysters sampled (Graph in Fig. 2). The annual pattern of larval production followed the same trends as those found for the proportion of the population brooding. A x~ goodness-of-fit test was used to further investigate the annual pat- tern of brooding. The data were pooled for each month, regardless of calendar year, because of the small proportion of oysters found brooding (Sokal and Rohlf 1987). The proportion of brooders varied with the time of year (Xn = 26.0, p < 0.01). In the months of September and October, considerably more brooding oysters were encountered than were expected, whereas m the months of January and July, there were fewer. To establish if different-sized animals were brooding at differ- ent times of the year, an analysis of variance was used to compare the mean size of brooding oysters for each month of the year (data pooled for 3 y ). The analysis of variance showed that there was no difference in the mean size of brooding oysters between months of the year (F|,|„ = 1.2, p > 0.05). An analysis of variance showed that there were differences in the number of larvae in broods from different months of the year (F, = 2.0, p < 0.05). The highest mean numbers of larvae per brood were found in the months of October. September, Au- gust, and November, whereas the lowest were found in January and May (Fig. 3). The water temperatures recorded at this site followed a general seasonal cycle, but with some variability that was probably due to the tidal nature of the harbour (Fig. 2). The highest temperature recorded was 24.3°C on January 10, 1994, and the lowest was ll.l°C on July 19, 1993. Annual Pattern of Brooding — Hauraki Gulf Twenty monthly collections of oysters were made between April 1994 and December 1995. In total, 1,548 adult oysters were sampled, of which 127 (8.2%) were found to be brooding. In addition. 19 brooding oysters were found among oysters remain- ing after the mam monthly sample had been processed. Data on the brood size of these oysters were included in the analyses below, where appropriate, in order to increase the sample size. Oysters were found to be brooding larvae throughout the period of the study because brooding oysters were encountered in each of the 20 monthly samples (Chart in Fig. 4). The highest proportions of brooders in all of the monthly samples were in December 1995 (I6.77f) and September 1994 (14.7%). The lowest proportions of brooders in all of the monthly samples were in April 1994 ( 1 .3%) and May 1995 (2.5%). There tended to be increased brooding activity in spring to early summer (September to December) and generally less brooding activity over the remainder of the year. The annual pattern of larval production for the population was estimated by dividing the total number of larvae found in each monthly sample by the total number of oysters sampled (Graph in Fig. 4). The annual pattern of larval production followed the same trends as those found for the proportion of the population brood- ing. A goodness-of-fit test was used to further investigate this an- nual pattern of brooding. The proportion of brooders encountered varied between the 20 mo sampled (x^q = 31.5, p < 0.05). In the months of December 1995 and September 1994, considerably more brooding oysters were encountered than were expected, whereas in the months of April 1994 and May 1995, there were fewer. To establish if different-sized animals were brooding at different times of the year, an analysis of variance was used to compare the mean sizes of brooding oysters for each of the monthly samples. The single brooding oyster found in April 1994 was excluded from the analysis. The analysis of variance showed that there was no difference in the mean size of brooding oysters throughout the months sampled (F,), i^f, = 1.68, p > 0.05). An analysis of variance showed that there were differences in the number of larvae per oyster throughout the months sampled *F|8 i->6 = 2.5, p < 0.005). The highest mean numbers of larvae per brood were found in the months of June 1994 and July 1994, whereas the lowest were found in March 1995 and January 1995 (Fig. 5). The water temperature in the Hauraki Gulf site followed a seasonal cycle similar to that of the Manukau Harbour, but had a slightly smaller range (Fig. 4). The lowest temperature was Annual Pattern of Brooding in Tiostrea chilensis 619 Figure 2. A chart showing the proportion of oysters in monthly samples from the Manukau Harbour that were brooding, with corresponding graphs showing water temperature and the monthly estimates of larval production (i.e., the total number of larvae in each monthly sample divided by the total number of oysters sampled). 13.0°C, recorded on both August 1 and August 14. 1995, and the highest was 23.1°C on February 17. 1995. DISCUSSION Our two populations of Tiostrea from northern New Zealand shared similar annual patterns of larval production. Both popula- tions were brooding larvae throughout the year, with less brooding activity in winter and increased larval production around spring and early summer. Larger broods of larvae appeared to be asso- ciated with increasing levels of brooding activity and lower water temperatures. Despite brooding year-round, the proportion of each population brooding during peak periods remained comparable to the highest proportions reported for populations elsewhere with shorter brooding seasons (Table 1). For both study populations, there were no differences in the size of oysters found brooding at different times of the year. The annual patterns of brooding in both of these populations 1 ^ f r I - I ^ I I I I I I I I I J I I I I I Monms of the Year {Data Pooled for Ttiree Yeare) Figure 3. A chart comparing the mean number of larvae for broods encountered in different months of the year in the Manukau Harbour (±standard error [S.Fl.]). 620 Jeffs et al. o 25 2 S ^ -5 Figure 4. A chart showing the proportion of oysters in monthly samples from the Hauraki Gulf that were brooding, with corresponding graphs showing water temperature and the monthly estimates of larval production (i.e., the total number of larvae in each monthly sample divided by the total number of oysters sampled). were very different from those previously reported for this species (Table 1). Earher studies have found that Tiosirea generally pro- duced larvae for at least 2 mo during spring to summer and some- times into autumn. Other than this study, Westerskov (1980) pro- vides the only account of winter brooding activity in this species. He recorded very low numbers of brooding oysters in Foveaux Strait over winter (0.01-0.1% during April and May and 0.2- 0.9% in July). These are much lower than for the corresponding periods during this study (see Figs. 2 and 4). Differences in the annual reproductive cycles of bivalves from geographically separated populations have been studied exten- sively, particularly among commercially important species such as oysters, mussels, and scallops (Gicse and Pearse 1974, Sastry 1979). Observed differences in breeding between populations have been variously attributed to genetic differences, food availability, latitude, and water temperature, or a combination of these factors (Newell et al. 1982, Barber et al. 1991, Wada et al. 1995). Cranfield and Michael (1989) observed significant differences rh rfl ih th [i rh A rh |f|iil||M|fii||l|l < 1 3 I I I t - < 1 3 1 1 Figure 5. A chart comparing the mean number of larvae for broods encountered in different months of the year in the Hauraki Gulf (±standard error IS.E.j). Annual Pattern of Brooding in Tiostrea chilensis 621 TABLE 1. Annual patterns of brooding recorded in populations of Chilean oysters from a range of locations. Location Annual Pattern of Brooding Comments Highest Proportion of Brooders Recorded Author(s) Auckland. N.Z. Wellington Harbour. NZ. Tasman Bay & Golden Bay. N.Z. Otago Harbour. N.Z. Foveaux Strait, N.Z. Foveaux Strait. N.Z. Pullinque. Chile Pullinque. Chile Bay of Concepcion. Chile Apiao. Chile Quempillen Estuary. Chile Quempillen Estuary, Chile Larvae released at least between January and May Larvae released from August to March Brooding at least from October to March Brooding September to May with a peak in November Brooding August to March with a peak in November to February; no brooders found April to July Brooding October to March with a peak in November Brooding September to February with a peak in December to January Larvae released November to February Larvae released November to March Larvae released December to March Larvae released December to January Larvae released end of October with a peak at end of December to early January Inferred season from larval settlement on a rocky shore No samples taken April to July Inferred from limited sampling No samples taken April to August No samples taken May to August 5.6% in December 4.1% in October 15-20% late October to mid-November tin two populations) 8.5% in October Just over 3% in November About 15% in December Booth 1983, Luckens 1976 Hollis 1963 Cranfield & Michael 1989, Tunbridge 1962 Westerskov 1980 Stead 1971 Cranfield & Allen 1977 Soli's 1967 Inculmar 1982 Aracena et al. 1976 Padilla et al. 1969 Toro 1982 Winter et al. 1984 in the breeding of Tiostrea populations from southern and central New Zealand and concluded that these were consistent with lati- tudinal trends identified in many other benthic marine invertebrate species by Thorson ( 1 950) . On the basis of these latitudinal trends, they hypothesized that breeding could be expected to start earlier and cease later in populations of Tiostrea at lower latitudes. This view is strongly supported by the results presented here from two populations in northern New Zealand. Unfortunately, a similar trend cannot be established from results reported for Chilean pop- ulations because of the limited extent of the studies undertaken over a latitudinal range. Water temperature, which usually varies with latitude in a moderately uniform manner, frequently has been assigned a dom- inant role in synchronizing reproductive cycles in marine inverte- brates (Newell et al. 1982). Indeed, Westerskov ( 1980) studied the timing of the production and release of larvae in Tiostrea in the Otago Harbour over 4 y and concluded that water temperature was critical in triggering the onset of gonad ripening and the subse- quent release of larvae. Gonads ripened over winter and into spring, but larvae were not incubated until the water temperature rose above 9-IO°C in late August/September. Subsequently, the water temperatures had always reached at least 14.5°C before the first spatfall and I5.9°C before the maximum spatfall was re- corded. These findings are consistent with the different water tem- peratures and annual patterns of brooding found in populations of oysters over a range of New Zealand latitudes. For example, water temperatures at our two northern populations never fell below IO°C, and hence, the incubation of larvae at both of these popu- lations continued uninterrupted throughout the year. Therefore, our results from low-latitude populations in New Zealand tend to confirm the importance of water temperature in regulating the annual cycle of reproduction in Tiostrea. This conclusion has implications for addressing the existing shortage of spat for the aquaculture of this species. In particular, there have been continuing difficulties in developing a hatchery technique for conditioning and synchronizing larval production in Tiostrea broodstock (Ramorino 1970. DiSalvo et al. 1983. Wilson et al. 1996). Our findings point toward the manipulation of water temperature as the most fruitful area for research aimed at improv- ing the hatchery production of larvae from broodstock. ACKNOWLEDGMENTS We thank Maria Buchanan for her translations of the Chilean research papers and the many people who helf)ed in the field and 622 Jeffs et al. laboratory work for this research. Oscar Chaparro from the Uni- versity of Austral, Valdivia. Chile, kindly provided additional information on Chilean research work. Mark Morrison, National Institute of Water and Atmosphere, was instrumental in locating the Hauraki Gulf study population. Barbara Hickey from the Auckland Regional Council assisted by providing access to water temperature records. We are indebted to Professor Tony Walsby. University of Bristol. United Kingdom, for making facilities avail- able for the preparation of the manuscript. This work was funded by Contract 402 with the New Zealand Foundation for Science, Research & Technology. LITERATURE CITED Aracena, O.. G. Tobella & M. T. Lopez. 1976. Cultivo de ostras iOslrea chilensis, Philippi, 1845) en Caleta Leandro. Bahia de Concepcion, Chile. Bol. Soc. Biol. Concep. 50:197-207. Barber, B. J., S. E. Ford & R. N. Wargo. 1991. Genetic variation in the timing of gonadal maturation and spawning of the eastern oyster. Cras- soslrea virginicti (Gmelin). Biol. Bull. 181:216-221. Beu, A. G. & P. A. Maxwell. 1990. Cenozoic mollusca of New Zealand. N.Z. Geol. Sun\ Pal. Bull. 58:336-341. Booth, J. D. 1983. Studies on twelve common bivalve larvae, with notes on bivalve spawning seasons in New Zealand. N.Z. J. Mar. Freslmar. Re.s. 17:231-265. Chanley, M. H. & P. Chanley. 1991. Cultivation of the Chilean oyster, Tiostrea chilensis (Philippi, 1845), pp. 145-151. /n; W. Menzel (ed.). Estuarine and Marine Bivalve Mollusk Culture. CRC Press, Inc , Boca Raton, Fl. Cranfield, H. J. & R. L. Allen. 1977. Fertility and larval production of oysters in an unexploited population of oysters Osirea liiiana Hulton, from Foveaux Strait. N.Z. J. Mar. Freshwal. Res. 11:239-253. Cranfield. H. J. & K. P. Michael. 1989. Larvae of the incubatory oyster Tiostrea chilensis (Bivalvia: Ostreidae) in the plankton of central and southern New Zealand. N.Z. J. Mar. Freshwai. Res. 23:51-60. DiSalvo, L. H., E. Alarcon & E. Martinez. 1983. Induced spat production from Osirea chilensis Philippi 1845 in mid-winter. Aqiiaciiltiire 30: 357-362. Giese, A. C. & J. S. Pearse. 1974. Introduction: general pnnciples. pp. 1-49. In: A. C. Giese and J. S. Pearse (eds.). Reproduction of Manne Invertebrates; Acoelomate and Pseudocoelomate Metazoans. Aca- demic Press, New York. HoUis, P. J. 1963. Some studies on the New Zealand oysters. Zoo. Pub. Victoria Univ. Wellington. 31:1-28. Inculmar Consultores Ltd. 1982. Situacion actual y altemativas de opti- mizacion de la captacion y produccion de semilla en el parque ostricola de Pullinque. Informe tinal Mimeografiado. 85 pp., 7 anexos. Jeffs. A. G. & R. G. Creese. 1996. Overview and bibliography of re- search on the Chilean oyster Tiostrea chilensis (Philippi. 1845) from New Zealand waters. J. Shellfish Res. 15:305-311. Luckens, P. A. 1976. Settlement and succession on rocky shores at Auck- land, North Island, New Zealand. N.Z. Ocean. Instil. Mem. 70:1-64. Millar, R. H. & P. J. Mollis. 1963. Abbreviated pelagic life of Chilean and New Zealand oysters. Nature 197:512-513 Newell. R. I. E., T. J. Hilbish, R. K. Koehn & C. J. Newell. 1982. Temporal variation in the reproductive cycle of Mytilus edulis L. (Bi- valvia, Mytilidae) from localities on the east coast of the United Slates. Biol. Bull. 162:299-310. Osorio, R. C. 1979. Moluscos marines de importancia economica en Chile. Biol. Pesq. Chile 11:3-47. Padilla, M., M. Mendez & F. Casanova. 1969. Observaciones sobre el comportamiento de la Ostrea chilensis en Apiao. Bol. Inst. Fom. Pesq Santiago 10:1-28. Ramorino, L. 1970. Estudios preliminares sobre la crianza de Ostrea chil- ensis en el laboratorio. Biol. Pesq. Chile 4:17-32. Roughley. T. C. 1929. Monoecious oysters. Nature 124:793. Sastry, A. N. 1979, Pelecypoda (excluding Ostreidae). pp. 113-292. In: A. C. Giese and J. S. Pearse (eds). Reproduction of Marine Inverte- brates. Molluscs: Pelecypods and lesser classes. Academic Press, New York. Sokal, R. R. & F. J, Rohlf. 1987. Introduction to Biostalistics. 2nd ed. Freeman. New York. 363 pp. Solis. I. F. 1967. Observaciones biologicas en ostras (Ostrea chilensis Philippi) de Pullinque. Biol. Pesq. Chile 2:51-82. Stead, D. H. 1971. Observations on the biology and ecology of the Foveaux Strait dredge oysler (Ostrea lutaria Mutton). N.Z. Fish. Tech. Rep. 68:1^9. Thorson, G. 1950. Reproductive and larval ecology of marine bottom invertebrates. Biol. Rev. Cambridge 25:1-45. Toro, J. E. 1982. Estudio de Osirea chilensis en Quempillen (Chiloe). Comunicacion libre presentada a las 11 Jomadas de Ciencias del Mar. Concepcion, Chile. 14 pp. Tunbndge, B. R. 1962. Occurrence and distnbution of the dredge oyster [Ostrea sinuata) in Tasman and Golden Bays. N.Z. Fish. Tech. Rep. 6:1^2, Wada, K. T., A. Komaru, Y. Ichmiura & H, Kurosaki. 1995. Spawning peak occurs dunng winter in the Japanese subtropical population of the pearl oyster. Pinctada fucata fucala (Gould. 1850). Aquaculture 133: 207-214. Walne. P. R. 1964. Observations on the fertility of the oyster [Osirea edulis). J. Mar. Biol. Assoc. U.K. 44:293-310, Westerskov, K. 1980. Aspects of the biology of the dredge oyster Osirea lutaria Mutton, 1873. Ph.D. Thesis. University of Otago, Dunedin. New Zealand. 192 pp. Wilson, J. A.. O. R Chaparro & R J. Thompson. 1996 The importance of broodstock nutrition on the viability of larvae and spat in the Chilean oyster OiJrfd chilensis. Aquaculture 139:63-75, Winter, J. E., J. E. Toro, J. M. Navan-o, G. S. Valenzuela & O. R. Chaparro. 1984. Recent developments, status, and prospects of mol- luscan aquaculture on the Pacific coast of South America. Aquaculture 39:95-134. Journal ,:l Slwllfish RcM'iinh. Vol. 15. No 3,623-626. 1996. SALINITY TOLERANCE OF THE CATARINA SCALLOP ARGOPECTEN VENTRICOSUS-CIRCULARIS (SOWERBY 11, 1842) GISELE SINGNORET-BRAILOVSKY,' ALFONSO N. MAEDA-MARTINFZ,' TEODORO REYNOSO-GRANADOS,- ERNESTO SOTO-GALERA,' PABLO MONSALVO-SPENCER,- AND GABRIELA VALLE-MEZA^ ' Uiiiversidad Aiitoiioimi Mclvopolitana-Xochimilco Calz. del Hiwso HOU Col. Villa Qiiu'iiid. Mexico. D.F. CP 04960 'Centra de Inveslii>(ieiones Biologicas del Noroesle S.C. La Pa: Baja California Sur Apdo. Postal 1 28. CP 23000. Mexico ABSTR.ACT We investigated the s.ilinity tolerance range of the adult eatarina scallop {Argopeclcn yt'iilncosus-circuluris) by ex- posing Iheni to increased and decreased salinity in steps of 5 ppt. each step lasting 24 h. from normal salinity (37 pptl up to 57 ppt and down to 17 ppt at a constant 28°C. Our results indicate that A. vemricosus-circularis is a perfect osmoconformer. Hemolymph osmolality followed the same trend as the changing external media throughout the salinity levels tested. Survival records indicate that the salinity tolerance is restricted to 27-47 ppt. kti WORDS: Salinity tolerance, osmoregulation, catanna scallop, survival INTRODUCTION Scallops arc bivalves of great economic impoilance. The eata- rina scallop (Argopeclen venlricosus-circuUuis) exists in large stocks in the bays of Baja California Sur. Mexico (Tripp 1985. Aurioles-Gamboa 1992). yielding up to 700 tons of adductor mus- cle in 1988 (Felix-Pico 1991. Felix-Pico et al, 1991) and 5.186 tons during 1989 and 1990 (Maeda-Martinezet al. 1993). Some of these stocks have been overfished (Baqueiro el al. 1981). and others are subject to heavy exploitation. Consequently, the culti- vation of this species is gaining impoilance (Vicencio and Singh 1988. Castro-Ortiz 1993). Spat production in the laboratory (Maeda-Martinez et al. 1989. Maeda-Martinez et al. 1995). field collection, and growout process (Maeda-Martinez and Ormart- Castro. 1995) in the bays of Baja California Sur are well docu- mented (Cacercs-Martinez et al. 1993). To determine the cultiva- tion range of the species in other protected areas, we need to know the tolerance of the species to salinity changes, because most of the coastal lagoons along the Pacific coast of Mexico are exposed to heavy rainfall and freshwater runoff. Others in Baja California Sur and Sonora become hypersaline during the year because ex- tremely high summer temperatures cause rapid evaporation and there are bays where water transport from the open ocean is re- stricted physically or there is limited tidal interchange (Garcia 1988). The bay scallop (Argopecten irradians. Lammarck. 1819) is the Atlantic species analogous to the eatarina scallop (Winter and Hamilton 1985). This species has been found in salinities as low as 10 ppt and as high as 38 ppt (Castagna and Chanley 1973, Barber 1985). Mercaldo and Rhodes ( 1982) demonstrated a certain capacity of the bay scallop to withstand reductions in ambient salinity. They found survival >60'7f could be obtained from scal- lops exposed to 15 ppt at 24°C for 48 h. Shumway ( 1977) earlier investigated the effects of lowered salinity on Chlamys opercii- laris, a species closely related to A. ventricosiis-circularis, finding that the former can withstand rather large decreases in salinity. The effect of hypersaline conditions on molluscs is not well known. Osmoregulation in molluscs has been the subject of many investigations (Robertson 1964. Avens and Sleigh 1965. McAlis- ter and Fisher 1968, Pierce 1970. Bedford 1971. Gilles 1972. Gilles 1974. Gilles 1975. Schoffeniels and Gilles 1972. Pierce and Greenberg 1973. Hoyaux et al. 1976. Shumway 1977. Burton 1983). From these, we can conclude that the hemolymph of most marine molluscs is close to seawater in osmotic pressure and ionic composition. This is achieved by intracellular isosmotic regula- tion, where free amino acids play a role as solutes. Shell closure in bivalves has complicated investigations into the ability of ma- rine molluscs to hyperosmoregulate actively at low salinities. Studies of osmoregulation in the eatarina scallop may provide information on this subject because the shells of A. ventricosus- circiihiris have a byssal notch that does not allow the animal to isolate itself completely from an external medium. Similar to all scallop species. A. ventrico.sus-circulari.^ opens its shells when disturbed or exposed to the air. This behavior is opposite that shown by other bivalves under the same circumstances. Shell closure in bivalves has also obscured investigations re- lated to the time required by the animal to reach equilibrium with the external medium. Crowe (19811 suggested that the regulation of cell volume by solute extrusion is. in bivalves, a long-term emergency process. In this article, the results of an investigation on osmotic regulation in A. ventricosus-circidaris as a function of environmental hyposaline and hypersaline changes are presented. MATERIALS AND METHODS The osmoregulatory capacity in the eatarina scallop was esti- mated by hemolymph osmotic concentration (HOC) measurements against the external medium osmotic concentration (EOC) and mortality rates. Adults of A. venlricosus-circularis (2.4 ± 0.9 cm shell length and 2.32 ± 0.9 cm shell height: n = 300) were collected from cultured populations at Rancho Bueno. Bahia Magdalena. Baja California Sur, Mexico (Fig. 1). The animals were maintained in 70-L plastic tanks in the laboratory, with run- ning filtered ( 10 |xm pore size) seawater at 28 ± TC and 37 ppt. 623 624 Signoret-Brailovsky et al. Magdalena Island Figure 1. Catarina scallop {A. ventricosiis-circularis) culture area. and were ted with 1.5 x 10^ cells/mL of a mixture o( Isochnsis gatbana. Chaetoceros calcitrans, and Teiraselmis suecica (6:3: 1 ). After 3 days, three groups of 50 animals were placed in individual 70-L tanks with filtered seawater at 28°C and 37 ppt to carry out the experiment (Fig. 2). The organisms were subjected to a change of 5 ppt/day above and below 37 ppt (treatments 1 and 2). The test animals were moved from a tank at one salinity to another at a different salinity to subject them to the salinity changes. The hy- posaline or hypersaline solutions were made by either diluting seawater with fresh water or by adding a measured amount of sea salt to seawater. One group was kept at 37 ppt and served as the control. All treatments were in duplicate. The osmolality of the media was adjusted with a hand refractometer (Area. Inc. ) reading the equivalent salinity in ppt (accuracy ±0. 1'/f ). The osmotic pres- sure of 30-|jlL samples of hemolymph (HOC) and external medium (HOC) was determined in triplicate for all groups, approximately every 6 h, with a micro-osmometer ("Osmette S"; Precision Sys- tems. Inc.). Hemolymph was obtained by direct puncture of the pericardic Figure 3. HOC and mortality of A . ventricosus-circularis as a function of a hyposaline environment. cavity with capillary pipettes, after the mantle cavity was blotted dry with a tissue. Simultaneously, the mortality rate was recorded. RESULTS Osmotic concentration measurements (Figs. 3 and 4) indicate that scallop hemolymph is isosmotic with the external medium at 28°C under laboratory conditions. This condition was tested sta- tistically by comparing the regression coefficients (Parker 1979) of internal and external osmotic concentration measurements. Results (Table I ) show that the regression coefficient (slope) of the exter- nal versus the internal osmotic concentration in all treatments was the same, at the p > 0.01 level. The osmotic hemolymph concentration of the organisms in the control group remained at 1.100 mOsm/L. At salinities of 17 and 57 ppt. the HOC values were 344 and 1 .830 mOsm/L, showing a decrease at low salinities (Fig. 3) and an increase at high salinities (Fig. 4). Thus, according to these results, the animals were os- moconforming, because the hemolymph osmotic pressure fol- lowed closely that of the external medium. A. ventricosus-circularis was able to withstand salinity varia- tions from 27 to 47 ppt. Within this range, mortality was low (6 and 8% at 27 and 47 ppt). At salinities of 22 and 52 ppt, mortality i 60 E 1.700 z ° 1.500 S 1.300 8 1.100- u 5 900 O X 700 1=00 ~ I 50 1 a z f /: < ~%r- "^\ 20 y ^ =:— ; 1 36 HOURS -HOC ^MORTALITY »EOC 80 g 60 < C o s 40 " Figure 2. Experimental design. Figure 4. HOC and mortality of A. ventricosus-circularis as a of a hypersaline environment. function Salinity Tolerance of Catarina Scallop 625 TABLE 1. Regression coefficient (slope) analysis, comparing the equations describing variations or external and hemolymph osmotic pressure against time in .4. lenlricosus-circularis under different osmolality treatments. Salinity Correlation of Osmotic Pressure and Time Regression Coefficient (Slope) Comparison of External and Internal Regression Coefficients Degrees of Freedom VI V2 Hyposaline EXT- INT Hypersaline EXT INT -5.58 -4.12 6.24 S Of) 0.93 0.93 091 0 40 20 20 0.66 1.43 ' EXT. external, INT. internal. increases to 50-60% over the 24-h holding period. Outside this range, mortality increases drastically, reaching 91 and 95% at 17 and 57 ppt (Figs. 3 and 4). These results show a certain ability of the species to withstand gradual salinity changes in both directions from the salinity level of normal seawater. DISCUSSION It has been shown, by many authors, that the mechanisms used by most marine molluscs to cope with external variations in sa- linity are intracellular isosmotic regulation and shell closing. Shell closing has imposed a limit on the estimation of the response time needed for the animal to adjust to the changing environment (Tet- telbach et al. 1985). In this work, the response of the scallop to salinity changes was recorded without interference, because A. venlricosus-circularis lacks the shell-closing protective mecha- nism. Within certain limits, most marine molluscs maintain the in- ternal medium isosmotic with the external medium The he- molymph osmolality of A. ventricosus-circuUins followed the same trend as the external medium. Because in our experiments, the changes in salinity were done in 5-ppt steps, it was possible to see that the adjustment of hemolymph to the external medium was a slow process and that it takes some time to reach the external level. Regardless of the hemolymph adjustment to the media tested, the salinity tolerance of the species is restricted to the range 27-47 ppt, as shown by mortality records. This is in contrast with the results of Mercaldo and Rhodes (1982), who obtained 100% sur- vival in A . irradiuns transferred directly from seawater to 1 8 ppt during a 50-h test at 24T. Our results suggest, because of the limited salinity tolerance range of the species, that the catarina scallop aquaculture industry IS restricted to areas with stable salinity conditions, such as the larger open bays of Baja California Sur. This is in agreement with the natural distribution of the species in Mexico. ACKNOWLEDGMENTS The authors express their gratitude to the Centre de Investiga- ciones Biologicas del Noroeste, La Paz. B.C.S.. where this work was done. This work was partially supported by CONACYT (Na- tional Council of Science and Technology). CIB. La Paz. and Universidad Autonoma Metropolitana-Xochimilco. Thanks also to Dr. Ellis Glazier, CIBNOR, for the editing of the English lan- guage manuscript. LITERATURE CITED Aurioles-Gamboa, D. 1992. Inshore-offshore movements of pelagic red crabs PIciiromodes plunipes ( Decapoda, Anomura. Galatheidae ) of the Pacific coast of Baja California Sur. Mexico. Crusiaceaiui 62:71-84. Avens. A. C. & M. A. Sleigh. 1965. Osmotic balance in gastropod mol- luscs. I. Some marine and littoral gastropods. Camp. Biiichem. Phys- iol. 16:121-141. Baqueiro. E., I. Pena & J. A. Masso. 1981. Analisis de una poblacion sobre explotada de Argopecten circularis (Sowerby, 1835), en la Ensenada de la Paz, B.C.S., Mexico. Ciencia Pesqiieru I.P.N. Me.\- ico 1:57-65. Barber. B. J. 1985. Growth and reproduction of the bay scallop Argopec- ten iradicms at Its southern distnbution limit. J . Exp. Mar. Biol. Ecol. Bedford. J. J. 1971. Osmoregulation in Melanopsis trifasciala. IV. The possible control of intracellular isosmotic regulation. Comp. Biochem. Physiol. 40A:1025-I058. Burton, R. F. 1983. Ionic regulation and water balance, pp. 291-352 In: A, S. M. Salleudin and K. M. Wilbur (eds.). The Mollusca. vol. 5, Physiology, part 2. Academic Press, New York. Caceres-Martinez. C J Chavez-Villalba cS: D. Ramirez-Flippini. 1993. I Encuentro Regional para el Cultivo Comerclal de Almeja Catarina [Argopeclen circularis). Guaymas. Sonora. Mexico, Ed. Instituto de Acuacultura, Gobiemo del Estado de Sonora. Castagna & Chanley. 1973. Salinity tolerance of some marine bivalves from inshore and estuarine environment in Virginia water on the west- em Atlantic Coast. Malacologia 12:47-96. Castro-Ortiz. J. L. 1993. Fishery and cultivation of the catarina clam in Baja California Sur. Mexico, Bullelin No. 25. CICIMAR-IPN. July 1993. 1 p. Crowe. J. H. 1981. Transport of exogenous substrate and cell volume regulation in bivalve molluscs. J Exp. Zuol. 215:363-370. Felix-Pico. E. 1991. Aspectos biologico-pesqueros de la almeja catarina Argopecten circularis (Sowerby, 1835) (Mollusca; Pelecypoda; Pec- tinidae) en Bahia Magdalena, B.C.S., Mexico. In: Memorias del I Congreso de Malacologia. Caracas, Venezuela. Felix-Pico, E. F., J. Amador- Buenrostro, G. Bojorquez-Verastica, R. Morales-Hernandez, J. Verdugo-Hemandez & G. Lopez-Garcia. 1991 . New record of pectinid spat (Argopecten circularis) on artificial col- 626 Signoret-Brailovsky et al. lectors in Bahias Concepcion and Magdalena. B.C.S-. Mexico, In: Memorias of the 8th International Pectinid Workshop, Cherbourg, France. 12 pp. Garcia, E. 1988. Modificaciones al Sistema de Clasitlcacion Cliniatica de Koppen SIGSA ed. Mexico 76 pp. Gilles, R. 1972. Osmoregulation in three molluscs: Acanihociiona dis- crepans (Brown), Glycymeris glycymeris (L.) and MytiUis ediilis (L.). Biol. Bull. Mar. Biol. Lab. Woods Hole 142:25-35. Gilles, R. 1974. Metabolisme de acides amines et controle du volume cellulaire. Arch. Int. Phvsiol. Biochim. 82:423-589. Giles, R. 1975. Mechanisms of iono and osmoregulation, pp. 259-347. In: O. Kinne (ed.). Marine Ecology, vol. II, part I. Wiley Interscience. New York. Hoyaux, J.. R. Gilles & C. Jeuniaux. 1976. Osmoregulation in molluscs of the intertidal zone. Comp. Biochem. Physiol. 53A:36I-365. Maeda-Martinez, A. N., P. Monsalvo-Spencer & T. Reynoso-Granados. 1989. Tecnologia para la produccion intensiva de semillas de almeja catarina (Argopecten circidaris). Internal Publication. Ceniro de In- vestigaciones Biologicas de Baja California Sur, Mexico. Maeda-Martinez, A. N., P, Monsalvo-Spencer & T. Reynoso-Granados 1995. Sistema para crianza intensiva en su etapa juvenil de alnieja catarina (Argopeclen circularis). Titulo de Patente No. 180212. Titular Centro de Investigaciones Biologicas del Noroeste, S.C. Maeda-Martinez, A. N. & P. Ormart-Castro. 1995. Sistema marino para crecimiento y engorda hasta la fase adulta de almeja catarina {Ar- gopeaen circularis). Titulo de Patente No. 18021 1 Titular Centro de investigaciones Biologicas del Noroeste, S.C. Maeda-Martinez, A. N., T. Reynoso-Granados, F. Solis-Marin, A. Leija-Tristan. D. Aurioles-Gamboa, C. Salinas-Zavala, P. Lluch-Cota, P. Ormart-Castro & E. Felix-Pico. 1993. A model to explain the for- mation of catarina scallop, Argopeclen circularis (Sowerby, 1835), beds, in Magdalena Bay, Mexico. Aquacuh. Fish. Mgml. 24:323-339. McAlister, R. O., & F. M. Fisher. 1968, Response of the limpet, Sipho- naria peainaia Linnaeus (Gastropoda. Pulmonata) to osmotic stress. Biol. Bull. Mar. Biol. Lab. Woods Hole 134:96-1 17. Mercaldo, R. S. & E. W. Rhodes. 1982. Influence of reduced salinity on the Atlantic Bay scallop Argopeclen irradians (Lammarck) at various temperatures. J. Shellfish Res. 2:177-181. Parker, J. 1979. Bioestadi'stica. Editorial Omega. 275 pp. Pierce, S. K., Jr. 1970. The water balance oi Modiulus (Mollusca Bi- valvia: Mytilidae): osmotic concentrations in changing salinities. Comp. Biochem. Physiol. 36:521-533. Pierce, S. K., Jr. & M. J. Greenberg. 1973. The initiation and control of free amino acid regulation of cell volume in salinity stresses marine bivalves. J. Exp. Biol. 59:435^40. Robertson, J. D. 1964. Osmotic and ionic regulation, pp. 283-308. In: K. M. Wilbur and C. M. Yonge (ed). Physiology of Molluscs, vol. I. Academic Press. New York. Schoffeniels, E. & E. Gilles. 1972. lonoregulation and osmo regulation in Mollusca. pp. 393-420. In: M. Florkin and B, Scheer (ed.). Chemical Zoology, vol. VII. Academic Press, New York. Shumway, S. E. 1977. Effect of salinity fluctuation on the osmotic pres- sure and Na, Ca, and Mg ion concentrations in the hemolymph of bivalve molluscs. Mar. Biol. 41:153-177. Tettelbach, S. T., P. J. Auster, E. W. Rhodes & J. C. Widman. 1985. A mass mortality of northern bay scallops Argopeclen irradians follow- ing a severe spring rainstorm. The Veliger 27:381-385. Tripp, A. 1985. Explotaciiin y cultivo de la almeja catarina (Argopeclen circularis) en Baja California Sur. MSc. Thesis. Centro Interdiscipli- nario de Ciencias del Mar. Instituto Politecnico Nacional. La Paz, B.C.S., Mexico. Vicencio, M. & J. Singh. 1988. Gui'a practica para el cultivo de almeja catarina. Secretaria de Pesca, Mexico. Winter, M. A. & P. V. Hamilton. 1985. Factors influencing swimming in bay scallops. Argopeclen irradians (Lammarck, I8I9). J. Exp. Mar. Biol. Ecol. 88:227-242. Journal of Shellfish Research. Vol. 15, No. 3. 627-634. 1996. ONTOGENETIC CHANGES IN OPTIMAL REARING TEMPERATURES FOR THE COMMERCIAL SCALLOP, PECTEN FUMATUS REEVE MICHAEL P. HEASMAN,* WAYNE A. O'CONNOR,' AND ALLEN W. J. FRAZER'^ 'NSW Fisheries Port Stephens Research Centre Salamander Bay. NSW. 2316. Australia 'MAF Fisheries South Private Bag 1926 Dunedin. New Zealand ABSTRACT Embryos, larvae, and early juvenile stages of the commercial scallop Pecienfiimaius Reeve, were held at temperatures ranging from 13 to 27°C. An incubation temperature of 18°C produced the greatest percent development of D-veligers from eggs. The growth rate of larvae increased from 2.5 ^.m/day at 15°C to a peak of 6.5 (j,m/day at 24°C but decreased with a further increase in temperature to 27°C. Age-specific larval survival decreased significantly with increasing temperature in the range 15-27°C. However, size-specitlc survival, which is a more meaningful measure of optimal reanng temperature, exhibited a pronounced peak value at an mtermediate temperature of 1\°C On the basis of these results, the maintenance of larval rearing temperatures between 18 and 21°C is likely to provide the ma.ximum yield of pediveligers. The growth of 3-wk-old spat (mean shell height. 1.04 ± 0.26 mml. held in the hatchery, increased from a negligible rate at 13°C to a maximum rate at 24°C. During the fifth and final week of the trial, a constraint to continued exponential growth became evident at all temperatures tested except I3°C. Survival and byssus attachment of spat were highest at temperatures supporting the highest growth rates. The use of byssus attachment as an indicator of favorable spat-growing conditions is discussed. Possible ecological implications of ontologic change in temperature optima are discussed in relation to variability in annual fisheries' catches. Occasions on which optimal temperature regimens for embryo development and larval and spat growth occur are rare in Jervis Bay. NSW. Aspects of El Nifio Southern Oscillation events and the oceanography of southeastern Australia are discussed as a possible mechanism by which such regimens might occur. KEY WORDS: Scallops, embryo, larvae, spat, temperature, survival, ecology INTRODUCTION Temperature is not the only environmental factor influencing growth of scallops; however, it is one of the more measurable and controllable parameters. It also has a profound influence on the survival and distribution of scallops (Nakanishi 1977. Ventilla 1982). including the commercial scallop Peaen fumatiis Reeve (Young and Martin 1989). Temperature directly influences the metabolic rate and survival of scallops (Nakanishi 1977. Ventilla 1982) and indirectly influences the nutritional environment (Wal- lace and Reinses 1985. Ito 1991). Accordingly, seasonal temper- ature regimens, critically influence the siting and seasonal timing of hatchery, farming, and stock enhancement programs. The geographical range of the Australian commercial scallop P.jumatus extends south from central New South Wales (NSW) to Victoria. Bass Strait, and southern Tasmania and west to the Gulf of St. Vincent in South Australia. It is found at depths ranging from 7 to 60 m on substrates varying from muddy sand to course sand (Young and Martin 1989). Mean monthly sea temperatures over the geographical range vary from minimum winter values of 9-1 1°C in southern Tasmania, to maximum summer values usu- ally in the range of 23-25°C on the central coast of NSW. As part of a research project to develop optimal hatchery and nursery rearing protocols for P. fumatus, embryos, larvae, and spat were exposed to a broad array of temperatures falling within the above-natural range and (presumably) within the tolerance lim- its of the species. This study complements research on the effects of temperature on reproductive conditioning, gamete storage, and fertilization in P. fumatus (Heasman et al. 1996a, Heasman et al. 1996b). MATERIALS AND METHODS The embryos, larvae, and spat used in this study were obtained from broodstock collected by divers in Jervis Bay, NSW, and road-freighted to the hatchery within 12 h of capture. Reproduc- tive conditioning and induced spawnings were conducted by the use of methods described by Heasman et al. (1996 a and b). In all cases, sperm and eggs, each from a minimum of five scallops, were used to reduce the effects of variable gamete viability. The seawater (35 g/kg salinity) used in all experiments was filtered to 1 |ji,m (pore size) and contained 1 mg/kg NaiEDTA (disodium ethylenediaminetetraacetic acid) as a precaution against metal contamination (Utting and Helm 1985). During Experiments 2 and 3. larvae and spat were fed mixed mieroalgal diets oi Pav- lova lutheri (Droop) Green, Tahitian Isochrysis aff. galbana Green (clone T. ISO), and Chaetoceros cakitrans (Paulsen) Ta- kano, originally shown to be suitable for rearing larval Sydney rock oyster [Saccostrea commercialis) larvae (Nell and O'Connor 1991). Within this study, growth is defined as an increase in shell length for larvae or height for spat, whereas survival is the number of live scallop larvae or spat at a particular time expressed as a percentage of the original stocking density. Experiment 1 . The Effect of Temperature on Embryo Development to D-Veliger Stage Eggs from five scallops were pooled in a 5-L glass beaker filled with seawater (21 ± 0.5°C, 35 g/L) and thoroughly mixed with a perforated polyvinyl chloride plunger. The concentration of eggs with in the beaker was estimated from the mean count of four 627 628 Heasman et al. replicate 1-mL aliquots sampled while mixing and examined at 40 X magnification on a Sedge wick rafter slide. Within an hour of spawning, sufficient sperm was then added to ensure that between one and five sperm were visible at the periphery of each egg (Heasman et al. 1996b). Eggs were again thoroughly mixed within 15 min of fertilization, and samples of a volume that was calcu- lated to contain 5,000 fertilized eggs were collected with an ad- justable automatic pipette and transferred to 1-L beakers filled with filtered seawater. Replicate sets of four 1-L beakers, stocked at five fertilized eggs/ml, were maintained at each of five temperatures — 16, 18, 21. 24, or 27°C (±0.5°C)— with waterbaths fitted with thermo- statically controlled immersion heaters. The water baths were housed in a coolroom held at a constant air temperature of 14 ± 1°C. After 48 h, the seawater in each beaker was thoroughly mixed as described above. A 20-iiiL sample, made up of four replicate 5.0-mL samples, was taken from each beaker, and the number of fully developed D-veligers was determined by dispersing the sam- ples on petri dishes and counting larvae with the aid of a dissecting microscope (40 x magnification). The total number of D-veligers in 20 ml was expressed as a percentage of the original number of eggs stocked in that volume (100 eggs). Experiment 2. Effect of Temperature on Larval Growth and Survival Fertilized eggs were stocked at approximately 50/mL into 90-L aerated polyethylene cylindroconical rearing vessels filled with filtered (1 |i.m pore size) seawater at 21°C. After 48 h, sufficient D-veliger larvae were collected to stock 20 90-L aerated cylindro- conical polyethylene tanks at 5/niL. Four replicate tanks were held at each of five temperatures ( 15. 18. 21, 24, or 27 (±0.5]°C) by supporting the tanks in 1 ,000-L temperature-controlled water baths (Fig. I). Every 48 h, the entire contents of each 90-L tank was siphoned onto a 45-p.m-pore-size nylon mesh screen that retained the larvae. Larvae from each 90-L tank were resuspended in 4 L of seawater, and a 1-mL sample was collected. From this sample, mean larval size (greatest shell length parallel to the hinge) and survival were determined with a dissecting microscope (40 x magnification) fit- ted with an eyepiece micrometer. The seawater in each 90-L tank was replaced with fresh, temperature-equilibrated, filtered ( 1 |xm pore size) seawater, and the larvae were returned to the tank. P.fumatus pediveligers normally begin to settlement behavior Airlift pump Immersion heater ^ 90 mm PVC downweller 150 um nylon screen 90 I cylindroconical tank 1 000 I water bath Figure 1. Mini downweller system used for growing P. fumalus spat. PVC, polyvinyl chloride. at shell lengths of 220 |j.m or more; however, some larvae as small as 1 85 |j.m had previously been found attached to the surface of the rearing tanks, thereby making samples taken from the water col- umn potentially unrepresentative. Consequently, survival data were recorded at the first appearance of 185-|jim larvae in each replicate. The effect of temperature on survival was calculated in terms of both age and size, with size-specific survival being con- sidered the most appropriate means for selecting optimal rearing temperature. Experiment 3. Effect of Temperature on Growth, Survival, and Byssal Attachment of Spat Hatchery-reared P.fumatus spat were maintained for 20 days after settlement on 150-(i,m-pore-size polyester mesh screens at 21°C in a downweller system (Bayes 1981). Forty spat (1.04 ± 0.26 mm. mean ± standard error) were randomly allocated to each of 20 miniature downweller systems (Fig. 1). Previously de- scnbed. 90-L cylindroconical rearing vessels fitted with lids were used to house individual miniature upwellers. These vessels were suspended in 1 .000-L water baths maintained within ±0.5°C of prescribed temperatures of 13. 17. 21. 24. or 27°C. Seawater was replaced with fresh, temperature-equilibrated seawater every 48 h. Salinities were monitored daily throughout the experiment to ensure that they did not vary outside the range of 35 ± 1 .0 g/L. The number of spat byssally attached to the internal walls and bottom mesh of each downweller screen was monitored. Whether or not scallops were byssally attached was determined by gently directing a stream of seawater from a squeeze bottle at individual spat 5. 20. and 40 min and 12 h after the initial stocking of the experiment and subsequently at weekly intervals. Each week, the miniature downwellers were placed in a 45 g/L hyper- saline solution prepared by the addition of artificial sea salt (In- stant Ocean; Aquarium Systems. Sarrebourg. France) to seawater. This procedure induced the rapid release of byssally attached scal- lops without imposing the traumatic injury and stress associated with mechanical methods of detaching spat (Heasman et al. 1994a). Detached spat were transferred to petri dishes, and the shell heights of live scallops were measured at 25 x with a dis- secting microscope fitted with a calibrated eyepiece micrometer. The total number of dead spat detected in each miniature upweller was recorded on each sampling occasion, their shells were re- moved, and the shell height was measured and recorded. RESULTS Experiment I . Effect of Incubation Temperature on Embryo Development to D-Veliger Stage The greatest mean yield of normal D-veliger larvae. 51 and 54%. respectively, occurred at the lowest test temperatures of 15 and I8°C (Fig. 2). The yield of D-veligers decreased sharply with increasing temperature above 18°C. falling to a mere \0% at 24°C and to 0% at 27°C. Experiment 2. Effect of Temperature on Larval Growth and Survival Larval growth (Fig. 3A), as indicated by mean growth incre- ment after 9 days (Fig. 3B). increased markedly with increasing temperature from 15 to 24°C but decreased with a further increase in temperature to 27°C. The age-specific survival of larvae (Fig. 4) was inversely related to rearing temperature, being highest at 15°C and lowest at 27°C. However, scrutiny of size-specific survival Rearing Temperatures for Pecten fumatus 629 2 CD 60 50- 40- ^ 30 20 10- 0 I 15 18 21 24 Temperature (°C) o — r- 27 Figure 2. Percent development of P. fumatus D-veligers after the in- cubation of emliryos at 15, 18, 21, 24, or 27°C. Values are means ± standard error (SE). data (Fig, 5) revealed a different pattern. Larvae survived to reach a shell height of 150 \x.m at all temperatures. At this time, size- specific survival was highest at higher temperatures, reaching a maximum value at 2rC (40% survival). It then decreased sharply and progressively with further increases in temperature to 24 and 27°C. Larvae in one replicate tank held at 24°C survived to a mean shell length of 166 (xm, but all died before the next water change. The experiment ceased when the largest larvae in a replicate com- menced settlement. Larvae only survived to metamorphosis in treatments held at 18 and 2rC. Experiment 3. Effect of Temperature on Growth, Survival, and Byssal Attachment of Spat P. fumatus spat with an initial shell height of 1 .04 ± 0.26 mm (mean ± standard deviation) grew exponentially at each of the five temperatures tested, i.e., 13, 17, 21, 24, and 27°C ± 0.5°C, during the first 4 wk of the trial (Fig. 6). Exponential equations were fitted to the first 4 wk of growth data (Table I) ahead of linear and other polynomial equations on the basis of higher r goodness of fit values (83.08-99.30%). This experiment was, however, terminated a week later when constraints to continued exponential growth became evident at all test temperatures other than I3°C. This growth "stalling"' in spat retained in the hatchery and fed diets selected for larval growth and survival was not un- expected, having occurred in all previous batches of P. fumatus spat reared at our hatchery. In all cases, spat growth ceased when they had either been retained in the hatchery beyond about 6 wk postsettlement or to a mean shell height in the range of 2—4 mm and fed a standard bivalve larval diet (Heasman et al. 1994b). Age-specific survival was lowest at the two lowest test tem- peratures, with survival at the end of the fifth week falling to 67% at I7°C and 60% at I3°C. By contrast, age-specific survival ex- ceeded 92% at the three higher test temperatures of 21, 24, and 27°C after 5 wk. Size-specific survival shows the advantage of maintaining spat above temperatures of 17°C (Fig. 7). 1/U 160 CO c o <) F 150 .r rn c 140 ik ■53 130 i9n 24°C . I 15''C J I l_ 0 3 6 9 12 15 18 Day (microns) 6.0 § c E 5.0 - 5 § o c 4.0 1 3.0 Q D 2.0 1 1 1 1 , 15 18 21 24 27 Temperature (°C) Figure 3. (A) Mean size of P. fumatus larvae held at 15, 18, 21, 24, or 27°C. (B) Mean daily growth increment (±SE) after 9 days of larval rearing. As indicated in Fig. 8. after stocking, most spat attached rap- idly to the bottom mesh or side walls of miniature downwellers screen. Peak numbers of spat attached were always attained within 12 h, and most were attained wilhm 40 min at the temperatures tested. Short-term percentages attached nevertheless increased with temperature, from a peak of about 77% at I3°C to over 95% at 24 and 27°C. Much greater differences in percentages of byssal attachment, however, developed over the 5-wk course of the ex- periment. The effects of rearing temperature on percentages of spat attached were thus consistent with its previously described effects on growth and survival. DISCUSSION Temperature has a marked effect on incubation and larval de- velopment in pectinids, beginning with the rate of cell division during early cleavage stages. For example, cell division rate in embryos is distinctly higher at 20°C than at 15 or IO°C (Zavarzeva 1981, cited in Cragg and Crisp 1991). Although we made no detailed observations of the effect of temperature on the incubation 630 Heasman et al. 100 80- ra 60 'E I 40 CO 20 27°C — r- 3 24*°C 21 °C I 9 Day — 1 1 1 12 15 18 Figure 4. Age-speciflc survival of P. fumatus larvae held at 15, 18, 21, 24, or 27°C. rate of P. fumatus. it was noted that development to the straight- hinge (D-veliger) first-feeding stage always occurred within 48 h when temperature was maintained at or above 18°C. The possi- bility that poor percent development to D-veliger stage was the product of increased bacterial levels was discounted on the basis of previous studies. The inclusion of several antibiotics, including erythromycin, oxolinic acid, and chloramphenicol, to experimen- tal batches of embryos failed to increase percentages developing to D-veliger at elevated temperatures (Heasman, O'Connor, and Frazer unpub. data). In a review, Cragg and Crisp ( 1991 ) found that time to meta- morphosis in pectinids is related to temperature and, when ex- 100 n 80 60 CO > ■^ =) w 15 40 -I CO _l 20 18°C 21 °C 27°C \24°C 110 120 130 140 150 160 170 Shell length (microns) Figure 5. Size-specific survival of P. fumatus larvae when held at a temperature of 15, 18, 21, 24, or 27°C. e'* 0) 3 0 CO 15°C 18°C 1 ,,0 24 °C 0" ,0' / A 6 27 °C / / .9 21 °C ■^/ ...r /,..y ■■■■'■ i--^ ^'"^ /^ /^ 13°C --:-:8^- .'---O'"" 0 1 2 3 4 5 Weeks Figure 6. Growth of P. fumatus spat held in the hatchery at 13, 17, 21, 24, or 27°C. pressed as an Arrhenius plot, is described by a single regression line. Figure 9 shows that equivalent data of P . fumatus. which ranges from 15 to 16 days at 18-19°C (Frankish et al. 1990) and 31 days at 13-15°C (Dix and Sjardin 1975), conform to this rela- tionship. These results for P. fumatus suggest possible benefits in further increases in rearing temperature, although larval growth rates have been shown to increase with increasing temperature to a maximum value and then to decline with a further rise in tem- perature. Ursin (1963) described the relationship between temper- ature and the time to complete a specified amount of growth as a symmetrical catemary curve: y = y^cosh p(x - x„) where y is time, x is the temperature at which development is most rapid, y^, is the development at .x„, and p is a temperature coeffi- cient. By the use of this relationship, the larval growth data of P. fumatus from this study were compared in Figure 10 with equiv- alent data for four other marine bivalves and a gastropod species compiled by Bayne (1983). Apices of the curves for each species coincide with their respective maximum growth rates. Maximum growth rate (up until a shell length of 150 \i.m) occurs about 24°C in the case of P. fumatus. This temperature corresponds with the highest sea temperatures that occur in late summer on the central coast of NSW (Wolf and Collins 1979), the northern extent of the TABLE 1. Exponential equations describing increases in shell height over 4 wk of P. fumatus spat reared at temperatures of 13, 17, 21, 24, or 27°C (n = 4). Temperature Exponential (T) Equation r^ 13 Y = g(0,0394+0.0559X) 83.86 17 Y — pi -0-0234 + 0. 1647X) 83.08 21 Y = g(0 0320 + 0.2231X) 96.70 24 Y = g(0 0348+0 3447X) 99.30 27 Y = gl -00394 + 0 3317X1 97.73 Rearing Temperatures for Pecten fumatus 631 100 80 to 60 ^ 3 to CO 40 Q. OT 20 13°C 0.5 1.5 2.5 3.5 4.5 5.5 Shell height (mm) Figure 7. Spat survival with shell height increase over 5 wk for P. fumatus spat held at 13, 17, 21, 24, or 27°C. Points indicate mean shell height and cumulative survival after each week. species range. However, size-specific survival data indicate tliat the maintenance of larvae in the hatchery at about 2 TC is likely to provide the maximum yield of larvae for settlement (Fig. 5). Al- though larvae initially grew rapidly at 24°C. they did not survive to metamorphosis. The reasons for this are unclear, but could involve the failure of larvae to maintain a net positive energy balance as a product of increased metabolic rates at higher tem- peratures. Overall larval survival m this experiment was poor, but was thought to reflect a general trend for poor yields from small experiment vessels in our hatchery. Unlike larvae, both the growth and the survival of P. fumatus spat were high at temperatures in the range of 21-27°C and great- 100 c a> E .c o (0 80 60 40 20 13°C OL I 1 I 2 3 Weeks Figure 8. Reattachment of P. fumatus spat in the first 12 h and in the succeeding weeks when held at 13, 17, 21, 24, or 27°C. In -2.5 -3.0 -3.5 -4.0 12 e • D Pecten fumatus • Other pectinids 20 33 55d DC • •bb • 21.1°C . I . . . 16.8 12.7 3.35 3.40 3.45 3.50 8.7 _i_J 3.55 T(K) 10 Figure 9. Arrhenius plot of the relationship between time from fertil- ization to metamorphosis and water temperature for various pectinids reared at close to the optimal temperature for each species (redrawn from Cragg and Crisp 1991 ). Datum points represented by squares are for P. fumatus using data from (a) this study and those of (b) Dix and Sjardin (1975) and (c) Frankish et al. (1990). Time is in days (d), temperature (T) is in degrees Celsius, and K is a rate constant. est at 24°C. up until the fifth week of the experiment. Comparisons of the growth of spat retained within a hatchery and fed diets of cultured algae with those placed in the wild or maintained in flow-through systems with natural phytoplankton have found the latter to be superior for both pearl oysters (Alagarswanii et al. 1989) and scallops (Bourne and Hodgson 1991). This was also the 0.10 0.08 1, 0.06 0.04 - 0.02 Mercenaria mercenaria Ostrea edulis Pecten fumatus Nassarius obsoletus 10 20 Temperature (°C) Figure 10. Growth rates of veliger larvae, calculated as the reciprocal of time in days from fertilization to pediveliger stage, related to tem- perature. Graphs are redrawn from Bayne (1983), with data for P. fumatus included. 632 Heasman et al. case for P. fumaius spat. The apparent barrier to continued growth, evident toward the end of this experiment, has been in- vestigated in recent studies (M. Heasman unpub. data 1995). The causative factor appears to be nutritional, with a shift in preferred diet from the larval diet used in this study toward diatoms at some point after metamorphosis. Regardless, the high cost of hatchery algal production encourages the deployment of spat to field nurs- ery systems, thereby avoiding growth retardation due to dietary factors. The preference exhibited by spat for elevated tempera- tures, then, raises the prospects of greenhoused nursery systems or perhaps farming this species as far north as Queensland, i.e. , up to 1,000 km north of its natural geographic range, at mean monthly sea temperatures as high as 27°C. However, field farming trials are needed to test this assertion. One means to quickly assess the suitability of more northern growout sites may be byssogenesis. In scallops, it has been pro- posed as a useful bioassay for potentially toxic substances (Roberts 1973, cited in Paul 1980a) and has been suggested to be a useful observation for the establishment of temperature optima in the scallop Chlamys opercularis (Paul 1980a). Byssogenesis by P. fumaius spat within 1-2 h of their transfer to new environments consistently reflected subsequent growth and survival under the particular set of physiochemical conditions involved. Short-term byssal attachment percentages may additionally serve as a useful quick indicator of suboptimal nursery and farming sites such as those subject to pollution, excessive turbidity, or suboptimal water chemistry due to coastal runoff. Most studies on the influence of temf)erature on scallops have focused on the tolerance limits of adult scallops (e.g.. Paul 1980a. Paul 1980b). with particular reference to ecological implications. However, sublethal temperature effects can also limit geographic range. For example, where metabolic rate increases with increas- ing temperature, greater energy acquisition via increased phyto- plankton clearance is also required to maintain a positive energy balance. Barber and Blake (1985) suggest that, in this way. ele- vated metabolic rate limits the southerly distribution of Argopecteii irradians irradians in the United States. On the other hand, inter- actions between temperature and other environmental variables such as dissolved oxygen may restrict the northerly range of the same species (Voyer 1992). The large disparities between optimal temperatures of approx- imately 15°C for gonadal development. 15-18°C for fertilization and incubation. 21°C for larval growth and survival, and 24°C for growth and survival of juvenile P. fumaius found in this and re- lated studies (Heasman et al. 1994b. Heasman et al. in press b) are of ecological interest. The significance of these findings needs to be considered in relation to the life cycle of wild populations of P. fumaius. Studies by Jacobs (1983). Fuentes (1994). and Heasman and O'Connor (unpub. data) have revealed multiple (three or four) annual peaks in gonadosomatic index in Jervis Bay P. fumaius. These peaks can occur at approximately 1- to 2-mo intervals over an 8- to 9-mo breeding season, beginning in midautumn (April) and ending in late spring (November). This penod coincides with mean monthly seawater temperatures in Jervis Bay in the range of 14-17°C (May et al. 1978. CSIRO 1994). These temperatures are similar to those experimentally determined to be suitable for gonad conditioning (15°C; Heasman et al. 1996a). The relationship between spawnings in P. fumaius and subse- quent settlement and recruitment is variable and is thought to be greatl, influenced by environmental conditions (Zacharin 1994). The incidences of such variation are common in the literature. Between October 1989 and October 1990 in Jervis Bay. Fuentes ( 1994) recorded four peaks in reproductive activity, indicative of spawning events, yet only two pulses of spat settlement were detected. Further, as also observed in Tasmania and Victoria (Hor- de and Cropp 1987, Sause et al. 1987), recorded spat settlements sometimes produced negligible subsequent recruitment to the fish- ery. Throughout its range, major spat settlements of P. fumaius have emanated largely from spawnings in spring and early sum- mer, when larvae encounter warm and rising sea temperatures. Although this observation is in general agreement with the find- ings of this study, the probability of P. fumaius larvae generated from spring or even early summer spawnings encountering optimal temperatures around 2rC would seem very low. However, it is conceivable that intermittent '"boom" catches, which occur every 10 y or so in the NSW P. fumaius fishery (Hamer and Jacobs 1987) and comparable widely fluctuating catches reported from Tasmania (Zacharin 1989, Young and Martin 1989) and Victoria (Gwyther 1989), may coincide with unusually large and rapid increases in sea temperature in spring or early summer. Such events could result in a mass synchronized spawning, especially if preceded by extended (3- to 6-wk) periods of low and stable sea temperatures and a high abundance of phytokplanktonic food, fa- vorable to rapid synchronous growth and the development of go- nads. Mass spawning at a time most conducive to subsequent high growth and the survival of larvae would in turn enhance the prob- ability of spat settlement sufficient to overwhelm natural predation and hence to generate high-level recruitment to the fishery. Some support for the above hypothesis that booms in P. fuma- ius fisheries are linked with favorable but unusual thermal se- quences is provided by oceanographic studies of southeastern Aus- tralia. The southeast coast of Australia, including Jervis Bay, is subject to two major interrelated influences, the warm "East Aus- tralian Current" (EAC), originating in the coral Sea (Fig. 1 1) and flowing south along the eastern seaboard, and a deeper, cooler, nutrient-rich upwelling comprising an Ekman boundary layer. The latter is generated by the overlying EAC and is driven tangentially from the continental slope toward the coast. Flow patterns of the EAC are often intense and highly variable. Between latitude 27°C (Tweed heads) and 32°S (Tuncurry/Forster) (Fig. 11). the flow often consists of strong southward currents near the edge of the shelf and equally strong northward currents further offshore. South of 32°S. the current degenerates into large, counterclockwise ed- dies. Each year on average, four to six of these warm meandering eddies progress as far south as southern Tasmania. They may affect any given section of the coast, especially dunng spring and summer (September to February), for periods of 4 wk to 4 mo (Boland 1979). Seawater exchange in Jervis Bay (Fig. 1 1 ) occurs mainly as a near-surface inflow on the southern side of the entrance in phase with a deeper outflow on the northern side. Flushing times for the bay were estimated by Holloway et al. (1992) to vary from 10 to 74 days, with a median of 21 days. Those authors also detected large pulses of cold (14-16°C) shelf water at a depth of 30 m beneath a surface layer of warm (20-24''C) seawater at the bay entrance during summer 1989/90. These cold, nutrient-rich sea- water intrusions persisted for periods of up to 3 wk at the entrance. One such dramatic intrusion of cold (15°C). nutrient-rich conti- nental slope water was driven into Jervis Bay by a near-shore, warm-core eddy of the EAC in late November 1992. This intrusion is thought to have caused a dramatic algal bloom (Cephyrocapsa oceaiuca) that turned the whole bay milky for a month (Blackburn Rearing Temperatures for Pecten fumatus 633 115°E O Pecten fumatus Figure II. Maps of Australia, southeast Australia, and Jervis Bay. depicting the distribution of P. fumatus, position and conformation of the EAC, places referred to in the test, and the broodstock collection site. and Cresswell 1993). This event lowered bottom temperatures within the bay from 18 to 16°C. but only fleetingly. Bottom tem- perature subsequently returned to 1 8°C by mid-January and rose to a peak of only 2rc by mid-February 1993 (M. Heasman et al. unpub. data). This was considerably lower than peak February temperatures of 23°C in 1989 (CSIRO 1994) and 1994 (M. Heas- man unpubl data) and of 25°C in 1991 (CSIRO 1994). The warm- water eddies of the EAC and associated cold, nutrient-rich upwell- ings regularly cause coastal phytoplankton blooms along the NSW (Tranter et al. 1986) and Tasmanian coasts (Harris et al. 1987) in spring and summer. These eddies and upwellings therefore exhibit the potential to occasionally create ideal conditions for mass spawnmg, high subsequent spatfall, and thence, high-level recruit- ment of scallops. Harris et al. (1988), in an analysis of Tasmanian scallop catches from the 1940s to the 1960s, found that years of high incidence of ""Zonal Westerly Winds" (ZWW), linked to an El Nifio Southern Oscillation cycle with a mean periodicity of 1 1 y, "'appear to favour good recruitment perhaps by a link between high productivity in high ZWW years, and increase in spawning and high larval survivorship." Similarly, high recruitment in the Shark Bay Amusiimt balloti fishery was usually associated with weak Leeuwin current activity in the winter months (Joll 1994, Joll and Caputi 1995). It was thought that the strong current ac- tivity may flush A . balloti larvae from embayments along the coast or may expose larvae to less favorable, high-temperature, low- nutrient-level seawater. ACKNOWLEDGMENTS The authors thank Dr. Geoff Allan, Dr. Stephen Battaglene, Mr. Duncan Worthington, Mr. D. Stewart Fielder, and the anon- ymous reviewers for valuable editorial comments and assistance during the preparation of the manuscript. This study was under- taken as part of a Fishing Industry Research and Development Trust Fund, grant 91/53. Alagarswami. K . S. Dharmaraj. A. Chellam & T. S. Veiayudhan. 1989. Larval and juvenile rearing of the black-lip pearl oyster, Pinctada magaritifera (Linnaeus). Aquaculture 76:43-56. Barber. B J & N J. Blake. 1985. Substrate catabolism related to repro- duction in the bay scallop Argopecten irradians. as determined by O/N and RQ indexes. Mar. Biol. 87:13-18. Bayes.J. C. 1981. Forced upwelling nurseries for oysters and clams using impounded water systems, pp. 73-83. /«; C. Claus, N. de Pauw. and E. Jaspers (eds.). Nursery Culturing of Bivalve Molluscs. European Mariculture Society Special Publication 7. Bredene. Belgium. Bayne. B. L. 1983. Physiological ecology of marine molluscan larvae, pp. 299-343 In: N. H. Verdonk et al. (eds.) The Mollusca. Vol. 3. Aca- demic Press. New York. Blackburn. S. 1. & G. Cresswell. 1993. A coccolithophond bloom in Jervis Bay, Australia. Ausl. J. Mar. Freshwater Res. 44:253-260. Boland, F. M. 1979. A time series of expendable bathythermograph sec- lions across the east Australian current. Aust. J . Mar. Freshwater Res. 30:303-13. Bourne. N. & C. A. Hodgson. 1991. Development of a viable nursery system for scallop culmre. pp. 273-280. In: S. E. Shumway and P. A. Sandifer (eds). An International Compendium of Scallop Biology and Culture The World Aquaculture Society, Baton Rouge. LA. Cragg, S. M & D J. Crisp. 1991. The biology of scallop larvae, pp. 75-122. In: S. E. Shumway (ed.). Scallops: Biology, Ecology and Aquaculture. Developments in Aquaculture and Fisheries Science, vol. 21. Elsevier. Amsterdam. CSIRO. 1994. Jervis Bay Baseline Studies. Final Report. 326 pp. CSIRO. Australia. LITERATURE CITED Dix. T. G. & M. J. Sjardin. 1975. Larvae of the commercial scallop Pecten meridionalis from Tasmania. Australia. Aust. J. Mar. Fresh- water Res. 26:109-112. Prankish. K.. L. Goard & W. A. O'Connor. 1990. Hatchery scallop cul- ture success in NSW. Austasia Aquaculture Mag. 4:10-1 1 . Fuentes. H. R. 1994. Population biology of the commercial scallop {Pecten fumatus) in Jervis Bay, NSW. Memoirs of the Queensland Museum 36:247-260. Gwyther, D. 1989. Yield assessment in the Port Phillip Bay scallop fish- ery, pp. 122-131. fn: M. L. C. Dredge. W F Zachann. and L. M. Joll (eds.). Proceedings of the Australasian Scallop Workshop. Hobart. Australia. Hamer. G. & N. Jacobs. 1987. The biology, fishery and management of the commercial scallop [Pecten fumatus) in Jervis Bay. New South Wales. Wetlands Ausl. 6:39-47. Hams. G. P.. P. Davies. M. Nunez & G. Meyers. 1988. Interannual variability in climate and fisheries in Tasmania. Nature 333:754-757. Hams. G . C. Nilsson, L. Clementson & D. Thomas. 1987. The water masses of the east coast of Tasmania: seasonal and interannual vari- ability and the influence on phytoplankton biomass and productivity. Ausl. J. Mar. Freshwater Res. 38:569-590. Heasman, M. P.. W. A. O'Connor & A. W. J. Frazer. 1994a. Detach- ment of commercial scallop. Pecten fumatus. spat from settlement substrates. Aquaculture 123:401-407. Heasman. M. P., W. A. O'Connor & A. W. J. Frazer. 1994b. Improved Hatchery and nursery rearing techniques for P . fumatus Reeve. Mem- oirs of the Queensland Museum 36:351-356, 634 Heasman et al. Heasnian. M. P., W. A. O'Connor & A. W. J. Frazer. 19%a. Temper- ature and nutrition as factors in conditioning broodstock of the com- mercial scallop Peclen fumalits Reeve. Aquacuhure 143:75-90. Heasman, M. P., W. A. O'Connor & A. W. J. Frazer. 1996b. Effects of fertilisation and incubation factors on the quality and yield of scallop, Peclen fuimitus Reeve, larvae. Acjuaciilnire Res. 27:505-513 Holloway. P. E.. G. Symonds & R. Nunes Vaz. 1992. Observations of circulation and exchange processes in Jervis Bay. New South Wales. Aust. J. Mar. Freshwater Res. 43:1487-1515. Hortle, M. E. & D. A. Cropp. 1987. Settlement of the commercial scal- lop, Peclen fumaius (Reeve) 1855, on artificial collectors in eastern Tasmania. Aquacuhure 66:79-95. Ito, H. 1991. Japan, pp. 517-569. In: S. E. Shumway (ed.). Scallops: Biology. Ecology and Aquacuhure. Developments in Aquacuhure and Fisheries Science, vol. 21. Elsevier, Amsterdam. Jacobs, N. 1983. The growth and reproductive biology of the scallop Peclen fumaius (albal in Jervis Bay, NSW and the hydrology of Jervis Bay. Hons. Thesis, University of NSW. 55 pp. Joll, L. M. 1994. Unusually high recruitment in the Shark Bay scallop (Amusium ballon) fishery. Memoirs of ibe Queensland Museum 36: 351-356. Joll, L. M. & N. Caputi, 1995. Geographic variation in the reproductive cycle of the saucer scallop, Amusium balloii (Bemardi, 1861) (Mol- lusca: Pectinidae), along the Western Australian coast. J . Mar. Fresh- waler Res. 46:779-792. May, v., A. J. Collins & L. C. Collet. 1978. A comparative study of epiphyte algal communities on two common genera of seagrass in eastern Australia. Aiisl. J. Ecol. 3:91-104. Nakanishi, I 1977. Studies of the effect of the environment on the heart rate of shellfishes. 1 . effect of temperature, salinity and hypoxia on the heart rate of scallops. Bull. Hokkaido Reg. Fish. Res. Lab. 42:65-73. Nell, J. A. & W. A. O'Connor. 1991. The evaluation of fresh algae and stored algal concentrates as a food source for Sydney rock oyster, Saccoslrea commerciatis (Iredalc and Roughley), larvae. Aquacuhure 99:277-284. Paul. J, D. 1980a. Upper temperature tolerance and the effects of temper- ature on byssus attachment in the queen scallop Chlamys opercularis (L ). J. E.xp. Mar. Biol. Ecol. 46:41-50. Paul, J. D. 1980b. Salinity-temperature relationships in the queen scallop Chlamys opercularis. Mar. Biol. 56:295-300. Sause, B. L.. D. Gwyther & D. Burgess. 1987, Larval settlement, juve- nile growth and the potential use of spatfall indices to predict recruit- ment of the scallop Peclen alba Tate in Port Phillip Bay, Victona, Australia. Fisheries Res. 6:81-92. Tranter, D. J., D. J. Carpenter & G. S. Leech. 1986. The coastal enrich- ment effect of the East Australian Current eddy field. Deep Sea Res. 33:1705-1728. Ursin, E. 1963. On the incorporation of temperature in the von Bertalnaffy growth equation. Medd. Dan. Fisk.-Havunders 4:1-16. Utting, S. D. & M. M. Helm. 1985. Improvement of seawater quality by physical and chemical pre-treatment in a bivalve hatchery. Aquacuhure 44:133-144. Ventilla. R. F. 1982. The scallop industry in Japan, Adv. Mar. Biol. 20:309-382. Voyer, R, A. 1992. Observations of the effect of dissolved oxygen and temperature on respiration rates of the bay scallop Argopeclen irradi- ans. Northeast GulfSci. 12:147-150. Wallace, J. C, & T. G. Reinses. 1985. The significance of various envi- ronmental parameters for growth of the Iceland scallop, Chlamys is- landica (Pectmidae), in hanging culture. Aquacuhure 44:229-242. Wolf, P. H. & A. J. Collins, 1979, Summary of daily temperature and salinity for major oyster bearing estuaries of New South Wales. NSW Dept. Agriculture, Division of Fisheries. Misc. Bull, 2, 107 pp. Young, P. C. & R. B. Martin. 1989. The scallop fishenes of Australia and their management. CRC Critical Rev. Aq. Sci. 1:615-638. Zacharin, W. F 1989. Scallop fisheries management: the tasmanina ex- perience, pp 1-11. In: M. L. C. Dredge, W. F. Zacharin, and L. M. Joll (eds, ). Proceedings of the Australasian Scallop Workshop, Hobart, Australia, Zacharin. W. F. 1994. Scallop fisheries in southern Australia: managing for stock recovery. Memoirs of the Queensland Museum 36:241- 246. J, mnml of Shetifish Research. Vol. 15, No. 3. 635-643. 19%. OPTIMUM CONCENTRATIONS OF ISOCHRYSIS GALBANA FOR GROWTH OF LARVAL AND JUVENILE BAY SCALLOPS, ARGOPECTEN IRRADIANS CONCENTRICUS (SAY) YANTIAN T. LU AND NORMAN J. BLAKE Dc'partmeiU of Marine Science University of South Florida St. Petersburg. Florida 33701 ABSTRACT Bay scallops from Homosassa. FL. were spawned at the Department of Manne Science, University of South Florida at St. Petersburg, FL. Total dry weight (DW) and ash free dry weight (AFDW) were determined for larvae and juveniles. DW and AFDW (in milligrams) increased with increasing shell length (L, in millimeters, for larvae) or shell height (H, in millimeters, for juveniles), according to the following allometric equations: DW = 0.0693L- ^" and AFDW = 0.0459L- '" in larvae and DW = 0.0715H-'"'''' and AFDW = O.OISSH-*"^ in juveniles. Growth was determined for larvae and juveniles at 25°C under six algal concentrations of Isochnsis galhana (1-30 cells/|i.L for larvae and 1-50 cells/(j.L for juveniles). Growth rates increased as algal concentration increased. The maximum larval growth rate ranged from 7-23 ^nvday. The development of eyespots and metamor- phosis started earlier at higher algal concentrations. The optimal algal concentration for growth was 20 cells/^.L for larvae and 10 cells/ [J.L for juveniles. In juveniles, the growth rate was higher and Increased with increasing body size. The relative growth rate (daily percent increase in AFDW) was higher in larvae than in juveniles and decreased with increasing body size in juveniles. KEY WORDS: Bay scallop, Argopecieii trracliam concentnciis. larvae, juveniles, growth. INTRODUCTION Growth is the most integrated response of organisms to changes in their suirounding environmental conditions (MacDonald and Thompson 1985), among which food availability is one of the most influential, and thus best studied, factors. Growth occurs when an animal is in positive energy balance, i.e., when assimi- lated energy exceeds the maintenance needs of the animal, and additional energy .is used to increase body weight. The relationship between growth and food availability has been an important re- search subject of physiologic energetics for a variety of bivalves, especially some commercially important species, e.g., mussels (Widdows 1978, Bayne and Worrall 1980), oysters ( Abdel-Hamid et al. 1992, Beiras and Camacho 1994), and scallops (MacDonald 1988, Hollett and Dabinett 1989), but such information on bay scallops is limited (Cahalan et al. 1989). The early life history of the bay scallop is characterized by a planktonic larval stage lasting 10-19 days, depending on environ- mental conditions (Sastry 1965, Castagna and Duggan 1971). At the end of the planktonic stage, larvae settle onto a substrate, typically scagrass blades, where they metamorphose to juveniles (Belding 1910, Gutsell 1930). Juveniles remain attached and grow to a shell height of about 20 mm before they detach and settle to the bottom to begin adult life. Existing data suggest that the high- est mortality occurs during the larval and juvenile stages and mor- tality declines with increasing animal size (Castagna and Duggan 1971). The growth and survival of larvae and young juveniles in an- imals with planktonic stages, like the bay scallop, determine the success of recruitment into adult populations. Yamamoto (1964) found that only 5-109^ of newly settled Patinopecten yessoensis juveniles survive their first 2 mo of life in Mutsu Bay. Paynter et al. (1993) reported that survival rates of Crassostrea virginica from pediveliger larvae to 5-mm spat averaged less than 5% at the Horn Point Hatchery of the University of Maryland. Considerable fluctuations in the abundance of bay scallop populations occur along the Florida Gulf Coast, and although the causes of such fluctuations are not fully clear, they may be related to reductions in early embryonic development and larval survival. Mortality during larval and juvenile stages can be lowered by fast growth. Unfortunately, detailed information is lacking on the physiology of the early growth of this species in relation to environmental conditions, particularly food availability. This study was designed to examine the growth of larval and juvenile bay scallops. Argopeclen irradians concenlricus. under various concentrations of the unicellular algae (Isochrysis galbana strain TISO. /. galbana has been shown to be an adequate food species for bivalve culture and has been used in many studies as food for bivalve larvae (e.g.. Castagna 1975. Peirson 1983. Sprung 1984. MacDonald 1988. Lu 1989. Zhang et al. 1991). MATERIALS AND METHODS Bay scallops collected from Homosassa. FL, were spawned at the Department of Marine Science. University of South Florida. Mature scallops were allowed to spawn in a 500-L fiberglass tank at 24-26°C and 25-28%f . Fertilized eggs were allowed to develop for 20-30 h. at which time they became D-shaped larvae. The larvae were then filtered onto a 35-|xm-pore-size screen and re- leased into fresh seawater. Larvae were cultured at a density of 4-8/mL and were fed daily with 10.000-30.000 cells/mL of/. galbana, which was grown in t72 median in 10-L plastic bags. Seawater was replaced every day in the amount of one-third of the total volume. As soon as larvae started to develop eyespots, black plastic Thalassiu mimics were added to the larval tank as substrate for larval settlement. The daily food ration for juveniles was in- creased gradually from 30.000 to 100.000 cells/mL of/, galbana. Weight Determinations Every 2-3 days during their planktonic life, larvae were sam- pled from a stocking tank, filtered onto a 35-|j.m-pore-size screen, and resuspended to 200 mL of filtered seawater. Larval density was determined by counting five subsamples of 2 mL each under a microscope. About 1.000-4. ()()() larvae were filtered onto a pre- combusted (475°C) and preweighed Whatman GF/C filter (punched to 7-mm diameter), rinsed with a 3% ammonium for- mate solution, and dried at 60°C for 48 h. Filters with larvae were weighed on an electronic microbalance to ± 1 p-g. They were then 635 636 Lu AND Blake combusted in a muffle furnace at 475°C for 5 h and reweighed for ash weight. Total ash free dry weight (AFDW) was obtained by subtracting total ash weight from total dry weight (DW). The weights of juveniles were measured in groups (for juve- niles <2 mm in shell height) or individually (for juveniles >2 mm in shell height). They were rinsed with a 3% ammonium formate solution and transferred to a precombusted and preweighed What- man GF/C filter (cut to one-eighth of its original size) with a pipette for sinall individuals or forceps for larger ones. Total DW. total ash weight, and total AFDW of juveniles were determined by use of the same procedure as described for larvae. To remove the soft parts for shell-weight determinations, lar- vae and small juveniles were killed in fresh water, and later, the animals were placed in beakers with seawater. The seawater in the beaker was changed every couple of days in the course of 1-2 wk until the soft tissue could be removed by microorganisms. For larger juveniles, shells were obtained by the removal of the soft body tissue with forceps. Growth Rales All experiments were carried out at 25°C, a good approxima- tion of the temperature in late September and early October in Homosassa (Barber and Blake 1983), where the scallop stock was collected. Larvae (48 h old) from a stockmg tank were placed in 2-L plastic beakers containing filtered (0.5 (xm pore size) seawater at an initial density of I individual/mL. /. galhana was added to the beakers to make a series of cultures, each containing one of the six algal concentrations: I, 2, 5. 10, 20, and 30 cells/fjiL. Larval cultures with no /. galbana were used as controls. All larval cul- tures were duplicated and were placed in a 25°C water bath. Gentle aeration was provided to the beakers to supply oxygen and to keep food particles in suspension. Experimental media were changed every day by filtering larvae onto a 35-|jim-pore-size screen and resuspending them in freshly prepared media. In addition, algal concentrations were monitored daily with a Coulter Counter and were adjusted if variation occurred by >I0% from the original levels (MacDonald 1988). Each day, a sample was taken from each culture with a pipette, and the shell lengths of 20 larvae were measured with a microscope fitted with an ocular micrometer. Experiments lasted until larval settlement occurred. Two-liter plastic beakers were used for experiments with juve- niles 0.5- and 1 .0-mm shell height, whereas 16-L glass tanks were used for juveniles 2-5.7 mm. Twenty animals were placed in each beaker or tank, except for the 5.7-mm size class, which was held at 10 animals per tank. Growth was determined at 1 , 5, 10, 20, 30, and 50 cells/|j,L of/, galhana. For juveniles larger than 2 mm, /. galhana stocks from 1 ,000-mL beakers were pumped with a Manostat cassette pump unit into the tanks continuously at a rate of 1,000 niL/day, to keep up with the reduction of algal concen- trations caused by juvenile feeding. These algal stock solutions were prepared freshly every day at concentrations ranging from 5 to 400 cells/p.L, which were determined by calculations from ju- venile ingestion rate (Lu 1996). Algal concentrations in juvenile culture media were also monitored daily with a Coulter Counter and adjusted as needed. Experiments with juveniles of all sizes were terminated on the 5th day. when juveniles were gently re- moved from experimental containers, collected in a Petri dish, and measured under a microscope. RESULTS Weight Determinations Body weight increased with increasing shell length (for larvae) or height (for juveniles), and the relationship can be described by the allomctric equation W = aH'', where W is body weight and H IS shell length or height. The fitted parameters a and b for various weight-size relationships are listed in Table I . Measured data and fitted curves are shown in Figure 1 . Total AFDW of larvae as a percentage of total DW was high, averaging about 37^3%. and it increased slightly as larvae grew (Fig. 2a); however, juveniles showed a dramatic drop in the f)er- centage of AFDW from about 30% for 0.4-mm juveniles to about 9% for 5-mm juveniles. The percentage of AFDW becomes con- stant at approximately 8% for juveniles >5 mm (Fig. 2b). Shell AFDW increased in relation to shell height. When ex- pressed as a percentage of shell weight, shell AFDW decreases from a mean value of 12.5% in larvae to 2% in 5-mm juveniles. It further decreased to about 1.2% in larger juveniles (Fig. 3). Growth of Lanae and Metamorphosis The shell growth of larvae at various algal concentrations is illustrated in Figure 4. Mean growth rate increased to its maximum at shell lengths between 140 and 150 \i.m and then decreased as shell length further increased. The maximum growth rate showed a strong positive correlation with algal concentration, being 7 \x.ml TABLE I. A. i. concentricus: fitted parameters (a and b) for allometric relationships between body weight (mg) and shell length (larvae, mm) or shell height (juveniles, mm) (W = aH''). Stage Parameter Total DW Total Ash Wt Total AFDW Shell Wt Shell Ash Wt Shell AFDW Larvae a 0.0693 0.0345 0.0315 0.1574 0.1644 0.0024 b 2.715 2.622 2.755 3.583 3.680 2.531 ^ 0.910 0.859 0.870 0.950 0.974 n' 33 a 33 10 10 10 Juveniles a 0.0715 0.0571 0.0138 0.0592 0.0567 0.0023 b 3.069 3.140 2.664 3.105 3.119 2.648 e 0.987 0.983 0.985 0.987 0.984 0.891 n 81 81 81 47 47 47 ' n, number of datum points evaluated. Optimal Concentrations of /. galbana 637 60% 130 150 SheO length (fxm) 130 150 SheD length (^m) 190 Sh»a hagfat (nun) A Total AFDW • Tout iky ucight Figure 1. ,4. ;. concentricus. Allometric relationship between body weight and shell length of larvae (a) and shell height of juveniles (b). Fitted parameters for the curves are listed in Table I. 40% 30% Q 20% 10% 0 5 10 15 Shefl height (mm) Figure 2. A. i. concentricus. Total AFDW as a percentage of total DW in larvae (a) and juveniles (b). Fitted line in panel a: AFDW% = 0.271 + 0.()00865L(|xni), r' = 0.13. Fitted curve in panel b: AFDW% = 0.204H(nini) "^^ at H < 5 mm; at H > 5 mm, AFDW% becomes constant at around 8%. day at 1 cell/|jLL and increasing to 23 fxm/day at 30 eells/p-L. The following logistic growth function was fitted to the measured data; L, {a-b) + c -ril'S) where L, is shell length at time t; r is growth rate; and a, b, and c are constants. The fitted parameters are shown in Table 2, and the fitted curves are shown in Figure 4. Two-way analysis of variance (ANOVA) demonstrate that the growth of larvae was significantly influenced by algal concentra- tion and larval age (day) (Table 3). Multiple range analysis on least significant differences between means showed that the growth of larvae was significantly different between any algal concentrations from 0 (control) to 20 cells/jj-L, whereas there was no difference statistically between 20 and 30 cells/|j,L (Table 4). The development of eyespots was highly synchronized in the cultures of 20 and 30 cells/ |j.L (Fig. 5). Eyespots were observed on the 10th day after fertilization, and over 95% of larvae developed eyespots a day later. In contrast, no eyed larvae were found in the 1 cell/jxL cultures on the I Ith day and eyed larvae composed only 18% of the larval population in the 2 cells/jiL cultures. At I cell/jiL, the lowest algal concentration tested, larvae could de- velop eyespots and eventually metamorphose, but there was a 2-day delay in development compared with those larvae cultured at 10, 20, and 30 cells/(iL. The average shell length of eyed larvae was also a function of algal concentration, with higher average shell length found at higher algal concentrations. The metamorphosis of larvae was first recorded 1 1 days after fertilization at the three higher algal concentrations, 10, 20, and 30 cells/ ^lL; 13 days after fertilization at 5 and 2 cells/fjiL; and 15 days after fertilization at I cell/|a.L, demonstrating a strong corre- lation between the rate of development and algal concentration. The same trend was observed in the number of successful meta- morphoses. A higher percentage of metamorphosis was observed as algal concentrations increased, although the rate of successful 638 Lu AND Blake 4 6 8 10 Shell height (mm) 12 14 AFDW -Fitted Figure 3. .4. /. concentricus. Shell AFDW of larvae and juveniles and its relative content (as a percentage of shell weight) versus shell height. Fitted curve: AFDW(mg) = 0.00158H(mm)- '*". metamorphosis was not documented quantitatively. Little meta- morphosis was observed at algal concentrations of 1 cell/|jLL. Growth of Juveniles The growth rate of juveniles as a function of algal concentra- tion is shown in Figure 6. The growth rate increased as algal concentrations increased to 10 cells/ |jlL and then became less de- pendent on algal concentrations. The growth rate increased as juveniles increased in size up to 3 mm in shell height, after which, growth rate showed little change with body size. Two-way ANOVA showed that the growth rate of juveniles was signifi- cantly affected by both algal concentration and body size (Table 3). Multiple range analysis shows that there was a significant difference in growth between algal concentrations of 1 . 5, and 10 cells/|xL (Table 4), whereas at >10 cells/|a.L, growth was not significantly different. Growth at a concentration of 1 cell/jjiL was not significantly different from that of controls. The above growth rates in shell size were converted to growth rates in AFDW, and the results are shown in Figure 7. Parameters fitted for the allometric curves of growth rate in AFDW against shell height are listed in Table 5. All graphs shown in Figure 7 are plotted on the same scale for easy comparison. It is obvious that growth rates in AFDW at I and 5 cells/(iL were significantly slower than that at >10 cells/(xL. At all algal concentrations, larger juveniles showed higher growth rates. The relative growth rate (daily percent increase in AFDW) was generally higher in larvae than in juveniles (Fig. 8) and decreased with increased shell height in juveniles. It was 2I^4'7f/day in 150-|jim larvae. 20-25%/day in 0.5-mm juveniles, and about 10%/ day in juveniles larger than 5 mm. DISCUSSION The growth curve of shell length against time for bivalve larvae is generally sigmoidal in shape (Loosanoff et al. 1951. Bayne 1965). Sigmoidal growth was observed for the larvae of Ar- gopecten irradUms irradians (Lu 1989) and for the larvae o{ A. i. concentricus in this study. The growth of the shell is characterized by an initial and a final phase of slow increment and a middle, rapid phase. During early larval development, either the digestive system is not fully developed or the velum does not reach its full capacity to capture food. As a result, larvae are unable to take full advantage of the external food resources available, and they may still need to partially use energy reserves from eggs. The combi- nation of low energy acquisition and high demand (Lu 1996) may lead to a negative energy balance at this stage. As larvae further develop, their ability to filter food particles increases (Lu 1996). Positive energy balance at this stage leads to increased growth rates. Shell growth slows down just before meta- morphosis, when larvae are at a critical stage to leave the plankton and become benthic. It is during the planktonic stage that larvae accumulate energy reserves for metamorphosis. The growth rate of larvae increased as /. galhami concentration increased. The optimal /. galhana concentration corresponding to optimal growth of bay scallop larvae was 20 cells/jxL. This value is comparable to the 30 cells/ (xL reported for the optimal growth of the larvae of the Japanese scallop Patinopecten yessoensis (Mac- Donald I9S8). In contrast, the algal concentrations required for the optimal growth by larvae of other bivalve species were 10-100 cells/fiL for Mytilus edulis (Bayne 1965. Jespersen and Olsen 1982. Sprung 1984). 20-400 cells/ |jiL for Ostrea edidis (Davis and Guillard 1958. Walne 1956. Wilson 1979. Beiras and Camacho 1994). 100 cells/fjiL for Crassoslrea gigas (Abdcl-Hamid et al. 1992). 25-325 for C. virginica (Rhodes and Landers 1973). and 50— 4(X) for Mercenaria mercenaria (Davis and Guillard 1958). The /. galhana concentration of 1 cell/|ji.l is close to the lower limit for normal scallop larval growth and development. At this concentration, only a few larvae reached metamorphosis and com- pleted larval development successfully. However, growth and de- velopment was considerably slower. Energy acquisition through feeding at such a low algal concentration was no more than res- piration loss (Lu 1996). Low energy reserves resulting from low food levels and high metabolic demand may be the key to the failure of metamorphosis of the majority of larvae at this low food concentration. In another study. Sprung ( 1984) found that M. edu- lis larvae failed to reach the pediveliger stage at 1 cell/fj-L of /. galhana, although larvae started growing. Once metamorphosis is complete, the juveniles grow much faster, possibly because of the high capacity of their gills to catch food particles. The optimal /. galhana concentration for growth was 10 cells/^,L for juveniles, lower than the 20 cells/ jjiL found for larvae. This difference could be the result of the high weight- specific metabolic rates of larvae (Lu 1996) because of energy used for swimming, and hence, larvae require a denser food con- centration. Another possibility is that juveniles had an increased ability to obtain food particles so as to saturate gut capacity at lower algal concentrations. No previous information is available on the weight-shell size relationships of larvae and juveniles of the bay scallop. The ex- ponents determined for allometric relationships between AFDW and size in this study are comparable with those found for two other bivalve species: the mussel M. edidis and the oyster O. edidis. Jespersen and Olsen (1982) reported values of b to be 3.49 and 2.42 for larvae and young postmetamorphic mussels, respec- tively (soft tissue DW-shell length), whereas Sprung ( 1984) gave a b value of 3.02 for mussel larvae (total AFDW-shell length). Slightly lower values of 2.50 (AFDW-shell length) and 2.64 (DW- shell length) were found for O. edidis larvae (Beiras and Camacho 1994). AFDW of bay scallop larvae was found to be 26.2-50.3% (mean, 38.6%) of the total DW (including shells). A high AFDW Optimal Concentrations of /. galbana 639 Control and 1 ceU /^l 200 180 100 2ceUs/^l 8 10 12 14 8 10 12 14 5ceUs/^l lOcells/^1 12 14 20ceUs/^l 30cells/^l 6 8 10 Days from fertilization 12 14 6 8 10 12 Days from fertilization Figure 4. A. i. concentriciis. Shell growth of larvae at various /. galbana concentrations. Pitted parameters for the logistic functions are listed in Table 2. content was also found in the larvae of M. ediilis, composing 18.3^8.2% (mean. 32.6%) of total DW (calculated from data of Sprung 1984). Such a high AFDW content indicates that larvae are well adapted to the planktonic lifestyle, because high organic con- tent increases the buoyancy of larvae and hence reduces the energy needed to maintain their position in the water column. High or- ganic content also represents a substantial energy reserve that can be used later for metamorphosis (Holland and Spencer 1973, Bart- lett 1979. Rodriguez et al. 1990. Lu 1996). Survival during metamorphosis and early benthic stages is of- ten low. Juvenile bay scallops have a 50-80% mortality before reaching 2-mm shell height (Castagnaand Duggan 1971). Juvenile scallops are vulnerable to predation by crabs, starfish, gastropods, and bottom-feeding fish (e.g., Medcof and Bourne 1964, Elner and Jamieson 1979, Lake et al. 1987). The high mortality of young juveniles can be offset by fast growing during this devel- opmental stage. The daily growth of 0.5-mm juveniles can be as high as 24% of their body weight in terms of AFDW. This is in TABLE 2. A. i. concentricus: fitted parameters (a, b, and c) for logistic growth functions of larvae at various /. galbana concentrations. Parameters 1 cell/fiL 2 cells/fiL 5 cells/fiL 10 cells/jjiL 20 cells/ fiL 30 cells/|iL a 82.47 112.34 I1X.92 106.74 100.49 102.02 b 25.64 40.43 56.17 49.80 42.80 45.97 c 94 85.3 78.5 90.2 97.7 98.0 r" 0.434 0.366 0.412 0.621 0.876 0 922 r. growth rate. TABLE 3. A. i. concentricus: ANOVA for growth of larvae and juveniles. Stage Source of Variation Sum of Squares d.f. Mean Sq. F Ratio Significance Level Larvae Juveniles Main effects Age (day) 1,434.487.0 9 159,387.4 1,265.0 0.0000 Cell Concn. 380.780.4 6 63,463.4 503.7 0.0000 Residual 264,209.5 2097 126.0 Total (corrected) 2,040.696.9 2112 Main effects Shell size 3 1566E9 7 4.5095E8 9.659.4 0.0000 Cell concn. 3.0061E7 6 5.0IO2E6 107.3 0.0000 Residual 47152149 1010 46685.3 Total (corrected! 3.2453E9 1023 TABLE 4. A. i. concentricus: multiple range analysis (95% Tukey HSD) for growth of larvae and juveniles by /. galbana concentrations. Larvae Juveniles Level Least-squares Homogeneous Least-squares Homogeneous (Cells/^L) Count Mean Groups Count Mean Groups 0 225 114.2 X 99 2,712.4 X 1 319 141.1 X 148 2,786.9 X 2 328 143.8 X — — — 5 330 149.4 X 152 2,993.7 X 10 332 156.2 X 152 3,122.0 X 20 310 158.9 X 161 3,132.8 X 30 269 161.3 X 154 3.205.5 X 50 — — — 158 3,180.0 XX 2(X) 10 11 Days after fertiUzation 2 3 4 Shell height (mm) -*- 1 c./|il -^ 2 c./mI -m- 5 c./(il 10 c./nl -^ 20 c./^l -■- 30 c./)il ■ lc/|li - 3c/m1 10 c./|J -*- 20 c/nl -•- 30 cJul -a- 50 c/Hl Figure 5. A. i. concentricus. Development of eyespot in larvae at var- Figure 6. A. i. concentricus. Shell growth rate of juveniles in relation ious /. galbana concentrations, c, cell. to /. galbana concentration, c, cell. Optimal Concentrations of /. galbana 641 IccU/hI SccUs/mI 10ceUs/)U 20ceUs/^l 30oeU3/(U SOcells/jU -S 120 y % /* ? 100 / Q / \ 80 / a. /A a 60 1 40 a/ Q ^^ cS 20 n __,,^^ 0 12 3 4 5 6 Figure 7. A. i. concenlhciis. Growth rate (in AFDW) of juveniles in relation to shell height at various /. galbana concentrations. Fitted parameters for the curves are listed in Table 5. Datum points in circles are excluded from regression. TABLE 5. A. I. concentricus: Fitted parameters (a and b) for allometric relationship between growth rate (mg of AFDW/day) and shell height (mm) of juveniles at various /. galbana concentrations. Parameter 1 cell/(iL 5 cells/|tL 10 cells/|xL 20 cells/inL 30 cells/^L 50 cells/jiL a 0.583 1.503 2.547 2.560 2.546 2.519 b 1.802 1.901 2.040 2.146 2.196 2.201 t" 0.922 0.944 0.934 0.950 0.958 0.981 Lu AND Blake 3 4 Shell height (mm) I -^r 1 c./^il -9- 5 c./|U -m- 10 c./nl -A- 20 c./m1 -•- 30 c/m1 -b- 50 c./h1 | Figure 8. A. i. concentricus. Relative growth rate of larvae and juve- niles in relation to shell size at various /. galbana concentrations. contrast to the 8% found for larger juveniles of >5 mm in shell height. The faster growth of shell relative to soft body parts in young juveniles represents another protective strategy to reduce predation. Total ash weight as a percentage of total DW increased from 71% in 0.5-mm juveniles to 92% in 5-mm juveniles. The organic content of the shell also decreased with increasing animal size. This agrees with observations that shell composition under- goes a transformation from low-calcium to high-calcium content when larvae metamorphose to juveniles (Merrill 1961) and that later juvenile shells are strengthened and thickened. ACKNOWLEDGMENT The authors thank Dr. Joseph Torres and Dr. Dan Marelli for their comments and suggestions on the manuscript. LITERATURE CITED Abdel-Hamid, M. E.. M. H. Mona & A. M. Khalil. 1992. Effects of temperature, food and food concentrations on the growth of the larvae and spat of the edible oyster Crassostrea gigas (Thunbergl. J Mar. Biol. Assoc. India 34:195-202. Barber, B. J. & N. J. Blake. 1983. Growth and reproduction of the bay scallop, Argopeclen irradians (Lamarck) at its southern distributional limit. J. E.xp. Mar. Biol. Ecul. 66:247-256. Bartlett, B. R. 1979. Biochemical changes in the Pacific oyster. Crassos- trea gigas (Thunberg) during larval development and metamorphosis. Proc. Nail. Shellfish Assoc. 69:202 (abstractl. Bayne, B. L. 1965. Growth and delay of metamorphosis of the larvae of Mytilus edulis (L.). Ophelia 2:419-t43. Bayne, B. L. & C. M. Worrall. 1980. Growth and production of mussels. Mytilus edulis from two populations. Mar. Ecol. Prog. Ser. 3:317-328, Belding. D. L. 1910. A report upon the scallop fishery of MassachuseUs. including the habits, life history of Pecien irradians, its rate of growth, and other factors of economic value. Special Report of The Common- weaUh of MassachuseUs Commission on Fisheries and Game, Boston, MA. 150 pp. Bieras, R. & A. P. Camacho. 1994. Influence of food concentration on the physiological energetics and growth of Osirea edulis larvae. Mar. Biol. 120:427-435. Cahalan,J. A.,S. E. Siddall&M. W. Luckenbach. 1989. Effects of flow velocity, food concentration and particle flux on growth rates of juve- nile bay scallops Argopeclen irradians. J. E.xp. Mar. Biol. Ecol. 129: 45-60. Castagna, M. 1975. Culture of the bay scallop, Argopeclen irradians in Virginia. Mar. Fish. Rev. 37:19-24, Castagna, M. & W. Duggan. 1971. Rearing the bay scallop, Aequipecten irradians. Proc. Natl. Shellfish Assoc. 61:80-85. Davis, H. C. & R. R. Guillard. 1958. Relative value of ten genera of microorganisms as food for oyster and clam larvae. Fish. Bull. Fish Wildl. Serv. U.S. 58:293-304, Bluer, R. W. & G. S. Jamieson. 1979. Predation of sea scallops, Pla- copecten magellanicus. by the rock crab Cancer irroratus and the American lobster, Homarus americanus . J. Fish. Res. Bd. Can. 36: 537-543. Gutsell, J. S. 1930. Natural history of the bay scallop. Bull. U.S. Bur Fish. 46:569-632. Holland, D, L. & B. E. Spencer. 1973. Biochemical changes in fed and starved oysters, Osirea edulis L. during larval development, meta- morphosis and early spat growth. J. Mar. Biol. Assoc. U.K. 53:287- 298. Hollen, J. & P. E. Dabinett, 1989. Effect of ration on growth and growth efficiency of spat of the giant scallop, Placopecten magellanicus (Gmelin). Aquacult. Assoc. Can. Symp. 89-3:71-73. Jespersen, H. & K. Olsen. 1982. Bioenergetics in veliger larvae of Mytilus edulis L. Ophelia 21:101-113. Lake, N. C H., M, B. Jones & J, D, Paul, 1987 Crab predation on scallop (Pecien ma.ximu.'^) and its implication for scallop cultivation, J. Mar. Biol. Assoc. U.K. 67:55-64, Lossanoff, V. L,, W. S. Miller & P, B, Smith, 1951, Growth and setting of larvae of Venus mercenaria in relation to temperature, J. Mar. Res. 10:59-81, Lu, Y, T, 1989. Effect of zinc on the growth and development of larvae of the bay scallop Argopeclen irradians. Chin. J. Oceanol. Limnol. 7: 318-326, Lu, Y, T, 1996, Physiological energetics of larvae and juveniles of the bay scallop .Argopeclen irradians concentricus (Say), Dissertation, Univer- sity of South Flonda, St, Petersburg, PL, 160 pp, Mac Donald, B, A, 1988, Physiological energetics of Japanese scallop /'a- linopecten yessoensis larvae. J. E.xp. Mar. Biol. Ecol. 120:155-170. MacDonald, B. A. & R. J. Thompson. 1985. Influence of temperature and food availability on the ecological energetics of the giant scallop Placopecten magellanicus. 1. Growth rates of shell and somatic tissue. Mar. Ecol. Prog. Ser. 25:279-294. Medcof, J. C, & N, Bourne, 1964, Causes of mortality of the sea scallop, Placopecten magellanicus. Proc. Natl. Shellfish Assoc. 53:33-50. Merrill, A. S. 1961. Shell morphology in the larval and postlarval stages of the sea scallop Placopecten magellanicus (Gmelin), Bidl. Mus. Comp. Zool. Harxard 125:3-20. Paynter, K. T,, C, Caudill, D. Meritt, S. Gallager & D. Walsh. 1993. Protein, carbohydrate and lipid levels associated with metamorphic success in larvae of the eastern oyster, Crassostrea virginica. J. Shell- fish Res. 12: p 134. Peirson, W. M. 1983. Utilization of eight algal species by the bay scallop, Argopeclen irradians concentricus (Say). J. E.xp. Mar. Biol. Ecol. 68:1-11, Rhodes, E, W. & W. S. Landers, 1973, Growth of oyster larvae, Cras- sostrea virginica. of various sizes in different concentrations of the chrys- ophyte Isochrysis galbana. Proc. Natl. Shellfish Assoc. 63:53-59. Rodriguez, J. L., F. J. Sedano, L. O. Garcia-Martin, A. Perez-Camacho Optimal Concentrations of /. galbana 643 & J. L. Sanchez. 1990. Energy metabolism of newly settled Oslrea edidi.s spat during metamorphosis. Mar. Biiil. 106:109-111. Sastry. A. N. 1965. The development and e.xtemal morphology of pelagic larval and post-larval stages of the bay scallop. Aequipeclen irradians conceiilricus Say. reared in the laboratory. Bull. Mar. Sci. 15:417- 435. Sprung, M. I9S4. Physiological energetics of mussel larvae (Myiilus edii- lis). II. Food uptake. Mar. Ecol. Prog. Ser. 17:295-305. Walne. P. R. 1956. Experimental rearing of the larvae of Oslrea edulis L. in the laboratory. Fish. Invest. Land. II 20:1-23 Wilson. J- H. 1979. Observations on the grazing rates and growth of Ostrea edulis L. larvae fed on algal cultures of different age. J. Exp. Mar. Biol. Ecol. 3S:187-199. Widdows. J. 1978. Combined effects of body size, food concentration and season on the physiology of MmiIus edulis. J. Mar. Biol. Assoc. U.K. 5X: 109-124. Yamamoto. G. 1964. Studies on the propagation of the scallop, Pati- nopecten yessnensis (Jayl. in Mutsu Bay. Jpn. Mar. Res. Prat. Assoc. 77 pp. (In Japanese). Zhang. P.. Y. He, X. Liu. J. Ma. S. Li & L. Qi. 1991. Introduction, spat-rearing and experimental culture of bay scallop. Argopeclen irra- dians Lamarck. Chin. J. Oceanol. I.imnol. 9:I2.V131. Journal of Shellfish Research. Vol. 15. No. 3, 645-651. 1996. GENOTYPE-DEPENDENT SPAWNING: EVIDENCE FROM A WILD POPULATION OF PECTEN JACOBAEVS (L.) (BIVALVIA: PECTINIDAE) C. RIOS, J. CANALES, and J. B. PENA Inslitulo de Acuicidtitra de Tone la Sal C.S.I.C. E-12595 Ribera de Cabanes Caslellon, Spain ABSTRACT In order to investigate whether there was a genetic basis for spawning asychrony in a wild population of the Medi- terranean scallop. Pecien jacohaeus (L.). we scored 15 protein loci during a 6-mo penod in which maturation and spawning were expected to occur. Multilocus heterozygosity was chosen as the vanable to be related to the timing of spawning. Our results indicate that the more heterozygous loci in individuals, the earlier they tend to spawn. Possible ways to reach asynchrony during the spawning period, either by an earlier start of maturation or by a shorter maturation time in heterozygotes. are proposed. In addition, the implications of this genotype-dependent spawning time for the genetic structure of the population are also discussed. As others have postulated, we propose that this dynamic could be. in addition to other acting forces, one of the possible explanations for the heterozygote deficiency often recorded in marine bivalves. KEY WORDS: Pecien jacohaeus. heterozygosity, spawning asynchrony. heterozygote deficiency, western Mediterranean, Spain INTRODUCTION In recent years, experimental studies have demonstrated the existence of positive relationships between multiple-locus het- erozygosity at electrophoretically detectable enzyme loci and fit- ness traits in many natural populations of marine bivalve species: oysters (Singh and Zouros 1981, Fujio 1982, Koehn and Shumway 1992, Zouros et al. 1983, Rodhouse and Gaffney 1984), mussels (Koehn and Gaffney 1984, Diehl and Koehn 1985, Rodhouse et al. 1986), clams (Green et al. 1983, Garton et al. 1984, Koehn et al. 1988), and scallops (Foltz and Zouros 1984). However, most fitness-related characteristics refer to growth and viability, and with a few exceptions (Rodhouse et al. 1986, Hilbish and Zim- merman 1988, Barber et al. 1991, Bricelj and Krause 1992), there is relatively little information about the genetic background of reproductive features. The scallop Pecten jacohaeus (L.) is a pectinid of Mediterra- nean distribution. In spite of the commercial interest, there are only a few studies concerning their reproductive features (Valli and Dovier 1977, Valli 1979. Castagnolo 1991. Mestre 1992) and genetic structure (Huelvan 1985, Pena et al. 1994. Rios et al. 1995) in some natural populations of this species. As is true of other members of the family, it is a functional hermaphrodite, having gonads with distinct male and female portions. The ga- metes are produced simultaneously; however, spawning is most protandric. Fertilization occurs externally, and larva is planktonic for 3-4 wk before settlement occurs. In this species, self- fertilization has not been investigated to date. However, in the congeneric Pecten maximus (L.) (Mason 1958, Gruffydd and Beaumont 1972, Beaumont and Budd 1983) and other hermaph- roditic pectinids (Castagna 1974, Ibarra et al. 1995), some labo- ratory studies have recorded self-fertilization, although the impor- tance of this process in the field remains unknown. During a 2-y period (October 1989 to April 1991), we studied the gametogenic cycle of a wild population of the scallop P. ja- cohaeus off of the coast of Castellon (E. Spain, W. Mediterra- nean). After a preliminary analysis of data from the first season (Mestre et al. 1990). several reproductive dynamics were noted: ( 1) a strong seasonal nature was observed; (2) the spawning period for the whole population was between late winter and early spring; (3) during the period before gamete release, a slight degree of asynchrony was observed among individuals with regard to go- nadal maturation; (4) this asynchrony became very evident during the spawning period, where a fraction of individuals of the pop- ulation had either spent, partially spent, or ripe gonads. In order to investigate whether there was a genetic basis un- derlying these differences, an experimental program was designed for the second season. According to our hypothesis that the genetic structure of individuals could affect the timing of reproduction, one should expect individuals showing different reproductive be- havior to display differences in some genetic traits in the period of maximal asynchrony. In earlier phases, where asynchrony is far from being evident, the genetic differences perhaps will not be as clear. We have chosen to observe multilocus heterozygosity, an in- tegrating measure, as a genetic trait. As discussed above, several studies have related this parameter to fitness (for a review, see Zouros 1987). MATERIALS AND METHODS One hundred twenty scallops of the species P . jacohaeus were collected between November 1990 and April 1991 from a natural population, located in front of the Oropesa coast (Castellon, E. Spain) (Fig. I ) at a depth of 65-75 m. Investigation of the eventual genetic influence on reproductive timing was made on the basis of pooling samples in three periods: (1) November to December, at the beginning of gonad maturation: (2) January to February, at the maximal gonad maturation; (3) March to April, at the spawning time. All individuals collected were adults of the 3-y age class, de- termined by the presence of the annual growth ring on the shell according to Gibson (1956). Each individual was measured, weighed, and dissected. As a maturation index, macroscopic ex- amination and microscopic examination of the gonads were made. The gonads were classified into three stages of maturity according to the following — a scheme modified from a scale proposed by Mason (1958) for P. maximus — (1) recovering (stages III and IV); (2) ripe (stages V, half full, and VI, full); (3) spent (stage VII, partially spent and spent). Sections of the adductor muscle were homogenized in an equal volume of homogenization buffer (Tris. 10 mM; EDTA, 1.27 645 646 RfOS ET AL. c — L > /' 0" K i» y ho- ^ peSiscola (i SPAIN V / MEDITERRANEAN / SEA y" ,J*IDROPes* 40-N •A CASTEUON 1 S / Smflai U 1" . , l» 1.0/ 0' |,. |« |,. indicate no subpopulation differentiation. The deviation from zero Fsrlk 1) Figure I. Location of the sampling area. mM; NADP. 0.03 mM; pH 6.8) and centrifuged at 4"C for 10 min at 5,000 ,?. The supernatant was used as the enzyme source. En- zymatic electrophoresis was performed on horizontal starch gels according to standard protocols detailed in Pasteur et al. (1987). For each individual, genotypes at 15 loci coding for 12 enzymes were determined. These enzymes included aspartate aminotrans- ferase (AAT: E.C. 2.6.1.1), a-glycerophosphate dehydrogenase (a-GPD; E.G. 1.1.1.8), glucose phosphate isomerase (GPl: EC. 5.3.1.9), isocitrate dehydrogenase (IDH; E.C. 1.1.1.42), leucine aminopeptidase (LAP; E.C. 3.4.11.1), malate dehydrogenase (MDH; E.G. 1.1.1.37), NADP-malate dehydrogenase (ME: E.C. 1.1.1.40). mannose phosphate isomerase (MPl; E.G. 5.3.1.8), octopine dehydrogenase (ODH; E.G. 1.5.1.11). 6-phosphoglu- conate dehydrogenase (6-PGD; E.G. 1 . 1 . 1 .43). phosphoglucomu- tase (PGM; E.G. 2.7.5.1). and superoxide dismutase (SOD; E.C. 1.15.1.1). All enzymes were resolved with a Tris-citrate (pH 8.0) buffer system, except a-GPD and LAP. which were run with a Tris-citrate discontinuous buffer system (pH 6.7 gel buffer; pH 6.3 electrode buffer). For enzymes with multiple isozymic forms, loci were num- bered sequentially starting with the most anodally migrating sys- tem. The different alleles at each polymorphic locus were assigned values that indicate their mobility relative to the most common allele, which was designated as "lOO." The loci Aat-2, Idh-1, Idh-2, and Mdh-1, which were com- pletely monomorphic, were discarded in further analyses. The locus Aat-1 showed poor resolution, and therefore, it was not interpretable. At each period sampled, allelic frequencies and deviation of heterozygosity (D) between heterozygosity expected under Hardy- Weinberg equilibrium (He) and observed (Ho) heterozygosity (D = |Ho-He]/He, where negative D values indicate a deficiency of heterozygous genotypes) (Selander 1970), were calculated for each locus. Conformity to Hardy-Weinberg equilibrium of distri- bution of genotype frequencies (genotype classes not being pooled) was tested by a Markov chain method that provides un- biased estimations of the exact p value (Guo and Thompson 1992) by the use of the computer program GENEPOP (Raymond and Rousset 1994). Between-penods differentiation was quantified by means of Wright's ( 1965) Fst statistic, modified by Nci (1977) and Wright (1978) for muhiple alleles, with the program BlOSYS-1 (Swof- ford and Selander 1989). Values of this statistic close to zero was tested for each locus by computing x" = 2N a statistic that follows a x' distribution with (k - 1) ■ (s - 1) degrees of freedom, where N is the total sample size, k is the number of alleles for the locus, and s is the number of subsamples (Workman and Niswander 1970). The statistical significance of mean ¥^-y for overall loci was tested by x~ = 2N • F^^^ with (n - 1 ) degrees of freedom, where N is the total sample size and n is the number of subsamples (Workman and Niswander 1970). The number of heterozygous loci per individual (multilocus heterozygosity) was used as a measure of heterozygosity. Each individual was heterozygous at zero to four loci; no individual was heterozygous at more than four loci. The effects of heterozygosity on asynchrony were tested for March to April data and for November to December data by linear- by-linear association exact tests with Monte Carlo accuracy to estimate the exact p value (Mehta and Patel 1995). Comparison of heterozygosity within the group of ripe individuals at three time periods was tested by a Kruskal-Wallis one-way analysis of vari- ance (Sokal and Rohlf 1981) and by a linear-by-linear association exact test with Monte Carlo accuracy to estimate the exact p value (Mehta and Patel 1995). RESULTS The number of scallops at each stage of maturity within the heterozygosity level, referred to each period, is summanzed in Table 1. Allelic frequencies, deviation of heterozygosity (D), and conformity to Hardy-Weinberg equilibrium at each locus, within scallop samples from the three different periods and pooling sam- ples, are shown in Table 2. The FsT values estimated individually for the 1 1 loci, and combmed with the set. are presented in Table 3. The results show low FsT values for most loci. The null hypothesis of homogeneity TABLE \. Number of individuals at each gonad stage for a given heterozygosity class within each period. Hetero- zygosity Gonadal Stage Recovering Ripe Spent Period Class (III-IV) (V-VI) (VII) November 10 December 0 1 2 — 1 3 8 — 2 1 10 — 3 0 4 — 4 0 0 — N" 5 24 — January to February 0 — 5 — 1 — 12 — 2 — 4 — 3 — 5 — 4 — 0 — N — 26 — March to April 0 — 8 11 1 — 12 12 2 — 2 13 3 — 2 4 4 — 0 1 N — 24 41 N. sample size. Genotype-Dependent Spawning in P. jacobaeus 647 TABLE 2. Allelic frequencies (P|), heterozygote deficiency (D), and conformity to Hardy-W'einberg equilibrium by exact p value, at the II loci in each period and pooling samples from the three periods. polymorphic November to December January to February March to April Total Period Sampled Locus Alleles P. D P Pi D P Pi D P 0.019 0.000 NT 0.015 -0.657 0.0006 0.013 -0.654 0.0000 0.9SI 0.954 0.950 0.000 0.031 0.038 0.115 -0.225 0.0025 0.092 -0.078 0.0525 0.100 -0.092 0.0000 0.038 0.038 0.046 0.558 0.754 0.654 0.135 0.023 0.071 0.154 0.085 0.125 0.000 0.008 0.004 0.058 0.074 I. 0000 0.000 -0.579 0.0002 0.017 -0.346 0.0018 0.885 0.900 0.896 0.058 0.092 0.083 0.000 0.008 0.004 0.000 — NT 0.023 0.016 1.0000 0.013 0.008 1.0000 1 .000 0.977 0.998 0.981 0.000 NT 0.985 0.008 1.0000 0.988 0.006 1.0000 0.019 0.000 0.004 0.000 0.015 0.008 0.058 -0.563 0.0001 0.069 -0.751 0.0000 0.087 -0.757 0.0000 0.731 0.823 0.775 0.173 0.108 0.129 0.038 0.000 0.008 0.038 -1. 000 0.0196 0.023 -0.557 0.0020 0.021 -0.658 0.0000 0.962 0.946 0.962 0.000 0.031 0.017 0.000 0.109 1.0000 0.000 -0.392 0.0046 0.004 -0.228 0.0290 0.885 0.815 0.817 0.115 0.185 0.179 0.000 0.000 NT 0.000 0.024 1.0000 0.004 0.022 1.0000 0.981 0.969 0.971 0.019 0.031 0.025 0.038 0.020 1.0000 0.038 0.048 1.0000 0.029 -0.077 0.1162 0.962 0.923 0.929 0.000 0.023 0.025 0.000 0.015 0.017 0.981 0.000 NT 0.954 -0.306 0.1135 0.950 -0.301 0.0242 0.019 0.046 0.050 N = 26 N = 65 N = 120 a-Gpd 105 0,000 100 0.914 95 0.086 Gpi 110 0.103 105 0.069 100 0.517 95 0.121 90 0.190 85 0.000 Lap 103 0.017 100 0.897 97 0.086 94 0.000 Mdh-2 110 0.000 100 1.000 Me-1 100 1.000 95 0.000 90 0.000 Me-2 105 0.115 100 0.707 95 0.138 90 0.000 Mpi 105 0.000 100 1.000 95 0.000 Odh 105 . 0.017 100 0.759 95 0.224 6-Pgd 105 0.017 100 0.966 95 0.017 Pgm 105 0.000 100 0.914 95 0.052 90 0.034 Sod 100 0.914 90 0.086 N" = 29 -0.785 0.0048 0.068 0.1319 -0.281 0.2489 NT" NT -0.926 0,0000 NT -0.094 0.6946 0.009 1.0000 -0.369 0.0214 -0.355 0.1707 ' N. number of individuaK ^cored for each locus. ' NT, not tested. in the allelic frequencies from the three periods considered was accepted in all loci examined except in the Gpi and o(-Gpd loci, which were found to show significant differences. However, the mean of Fj^ for overall loci shows a value not significantly dif- ferent from zero (Fs^^ = 0.017, x" = 4.08. d.f. = 2, p > 0.1). It could be concluded that all scallops taken within each period were members of the same breeding population, regardless of the period sampled. During the spawning period (March to April), spent and ripe individuals showed differences in multilocus heterozygosity (Fig, 2). but these were not significant (linear-by-linear association; p = 0.0791 . one-tail). According to these results, although the more homozygous individuals tend to correspond with the delayed group (ripe individuals), whereas the more heterozygous individ- uals were largely confined to the spent group, because the test for linearity gave a not significant value, the causal relationship be- tween the variable multilocus heterozygosity vs. gonadal stage remains somehow unclear. Therefore, these results should be taken with caution. In the same way, during the November to December period, the recovering and ripe individuals show slight differences (Fig. 2), although these were not significant (linear- by-linear association; p = 0.0993. one-tail). Nevertheless, the average heterozygosity of ripe individuals from the three periods considered decreased with time (November to December, x = 1 .667; January to February, x = 1 .346; March to ApnI. X = 0.917; Kruskal-Wallis x^ = 8.368, d.f. = 2. p = 0.0152). Moreover, the linear association of the variables, multi- locus heterozygosity vs. time period, was highly significant (lin- ear-by-linear association; p = 0.0048. one-tail) (see Fig. 3). Therefore, these results confirm the hypothesis that animals rip- 648 RfOS ET AL. TABLE 3. Single-locus and multilocus values of F^y among scallops sampled within each period. Locus FsT X^ d.f. a-Gpd 0.022 10. 36-' 4 Gpi 0.026 31.20'' 10 Lap 0.004 2.88 ns' 6 Mdh-2 0.016 3.84 ns 2 Me-1 0.009 4.32 ns 4 Me-2 0.013 9.36 ns 6 Mpi 0.016 7.68 ns 4 Odh 0.016 7.68 ns 4 6-Pgd 0.003 1.44 ns 4 Pgm 0.011 7.92 ns 6 Sod 0.016 3.84 ns 2 Multilocus 0.017 4.08 ns 2 "p < 0.05. "p < 0.005. ' ns, not significant. ening earlier tend to be more heterozygous. This may represent a genetic control over spawning asynchrony. DISCUSSION As we postulated, individuals in different stages of gonad de- velopment during a given period show differences in genetic traits. The results indicate that the more heterozygous loci present in individuals, the earlier they tend to spawn. This became evident during the spawning period (see Fig. 2). In order for individuals with more heterozygous loci to spawn earlier, we hypothesize that three events could occur: ( 1 ) heterozygous individuals start to mature before; (2) even if a population was synchronous at the beginning of maturation, less time would be spent on gonad mat- uration in heterozygotes; (3) a combination of the two mechanisms could be involved, i.e., heterozygous individuals tend to start maturation earlier, but at the same time, they spend less time to reach spawning. These three hypotheses and their implications are summarized in Figure 4. Because under the three hypotheses, the final result is the same, our results from the third period (Fig. 2) do not provide evidence enabling a choice among these three possibilities. How- ever, the analysis of November to December data among recov- ering and ripe individuals and their degree of heterozygosity show slight, although not significant, differences. Our results on this last point agree with the third model: a combination of the two mech- anisms, earlier start and shorter time of maturation, seems to lead heterozygotes to an earlier spawn. This could also explain the results shown in Figure 3. It could be concluded that, even though a tendency for an earlier start time of maturation in heterozygotes seems to exist, our results do not provide clear evidence to support the idea that a nonsynchronous start of maturation is the only reason for the ev- ident asynchrony reached at spawning time. In this sense, these results serve as a pilot data set that might aid in the design of future, more definitive experiments. At present, there is no consensus on the genetic mechanism underlying the correlation of heterozygosity with phenotypic traits related to fitness (growth, viability, fecundity, etc.), and research- ers are divided into two principal points of view: (I) one school of thought supports the idea that the enzymes recorded by electro- phoresis are the causal agents of the correlation, i.e.. heterozy- gosity for these enzymes directly affects those traits (Koehn et al. 1988); (2) another school (Zouros et al. 1988) advances the view that the enzyme variants act as mere markers of genetic abnormal- ities that are responsible for the genetic variation in traits related to fitness, but cannot be detected by the electrophoretic assay in- volved. Thus, an individual that is multiply heterozygous for the marker genes will have a much lower probability of carrying one or more of these abnormalities compared with an individual that is multiply homozygous for the marker genes. On the other hand, researchers appear to be in agreement concerning the physiologic basis of this correlation. Some studies demonstrate that the effect of increasing heterozygosity is to endow an individual with a lower energetic cost of maintenance metabolism (Koehn and Shumway 1982. Carton et al. 1984, Hawkins et al. 1986. 1989. Koehn et al. 1988). The emerging physiologic interpretation of the correlation is that the increased level of heterozygosity enables an individual to sustain its basal metabolism with lower expenditures of ATP. Therefore, according to Koehn ( 1990), this would allow a higher allocation of energy for growth and reproduction. As a result, time of maturation would be shorter. However, it is imperative to note that the eventual effect of heterozygosity would be to increase production and not necessarily to induce changes in the allocation of energy. In this context, Rodhouse et al. (1986) concluded that, in the case of Mytilus ediilis. it is in older individuals that the majority of energy production is allocated for reproduction instead of growth. Therefore, it is later in the life of the organism when a higher production rate hypothctically associated with multilocus heterozygosity could be translated into increased gamete produc- tion. On the other hand, in younger individuals, the majority of energy production would be allocated for growth, and subse- November-December Recovering individuals MLH 2 MLH 0 20% ■'"'^'^^^20% Ripe individuals March-April Spent individuals Ripe individuals Figure 2. Proportion of individuals at multilocus heterozygosity (MLH) within each gonad stage during the November to December and March to April periods. Genotype-Dependent Spawning in P. jacobaeus 649 100] 20 I I nMLH3 BMLH2 ■ MLH1 ■ MLHO nov-dec jan-feb mar-apr Figure 3. Proportion of ripe individuals al each multilocus heterozy- gosity (MLH) within each of the three time periods. quently, the effect of heterozygosity on reproduction would not be detected. As far as we i^now. this is the first work that provides empirical evidence for a relationship between multilocus heterozygosity and the timing of reproduction in marine bivalves. Although Zouros et al. (19881 attempted to prove this relationship in a natural popu- lation of mussels, the authors failed to provide experimental evi- dence. This could be due to several reasons. In this sense. Koehn (1990) argued that the reproductive cycle of that population of bivalves had not been previously studied, and consequently, the period when spawning was expected to occur was unknown. In addition, the period considered ( I mol was too short for a degree of asynchrony to become evident. On the other hand, the animals were originating from a 2-y-old cohort, and as discussed above, the allocation of energy for reproduction appears to be age depen- dent (Rodhouse et al. 1986). According to Koehn (1990). this would be a critical age for the energetic balance. We tried to overcome these difficulties both by performing a preliminary study on the ganietogcnic cycle (Mestre et al. 1990) and by choosing animals old enough so that energy would be mostly allocated for reproduction. CONSEQUENCES OF THE SPAWNING PATTERN FOR THE GENETIC STRUCTURE OF THE POPULATION Until now. we have been dealing with the phenomenon of genotype-dependent spawning itself and the mechanisms that could be involved in its maintenance. However, at the same time, and because this dynamic leads to a nonpanmictic mating, it be- comes evident that it could have consequences on the genetic structure of the population. Thorough analysis of these consequences is far beyond the limits of this work. Nevertheless, we would like to add some notes on this matter, because genotype-dependent spawning has been regarded as one of the possible explanations for the most charac- teristic genetic trend among marine bivalve populations: the het- erozygote deficiency (Zouros and Foltz 1984, Alvarez et al. 1989). As proved by many experimental studies, this phenomenon seems to be a general trend among populations of marine bivalves. This is apparent for hermaphroditic and gonochoric species, for wild and cultivated populations, and also for different families of bivalves (see Zouros and Foltz 1984 for a review). Neither pec- tinids (Wilkins 1978. Beaumont and Beveridge 1984; Foltz and Zouros 1984; Gosling and Bumell 1988) nor other populations of P . jacobcieus (Huclvan 1985). including this particular population (Rios et al. 1995). appear to be an exception to this rule. Several studies attempted to determine the cause of this het- erozygote deficiency in marine bivalve populations (Zouros and Foltz 1984. Gaffney et al. 1990. Beaumont 1991). Within this context, the plausibility of the genotype-dependent spawning mod- els and subsequent nonpanmictic mating was first invoked by Zouros and Foltz (1984). Those authors considered a theoretical model under which a population spawns in two discrete periods of time, the union of gametes being random within each period. The probability of an individual to spawn in one of the two periods (or alternatively, the fraction of gametes released in each period) was made to be a function of its genotype. One of the one-locus dial- lelic models that the authors built and thoroughly explored was an ■"overdominance" model. It appears that under this model: (I) the heterozygote deficiency (£>) generated is a function of allelic fre- quencies (p): (2) it also depends on k. the fraction of a given genetic class that spawns in a different period; (3) for some com- binations of these parameters, high D values can be generated. Because this model assumes no changes on allelic frequencies over generations but changes on genotype frequencies, theoretically at least, it appears to give rise to a stable situation. Although the discussed models were one locus, its eventual prolongation to a multilocus situation could be supported on the context of the hypothesis termed "associative overdominance" (see Zouros and Mallet 1989), which relies heavily on the genetic structure of the population as well as that of the individual. Ac- cording to this, the fitness of individuals scored homozygous for several loci would be reduced compared with multiple-locus het- crozygotes as a result of homozygosity of deleterious genes at loci in linkage disequilibrium (Zouros and Mallet 1989). Moreover, the negative effects of deleterious loci "will be amplified by any process that causes an excess of homozygosity in the population" (Zouros and Mallet 1989. p. 319). such as inbreeding. Neverthe- less, that effects will be present to some degree when sampling multiple-locus homozygotes in outbred populations (Beaumont 1991). However, a multilocular extension of the model of Zouros has been believed to be unlikely (Gaffney et al. 1990) to generate large, single-locus, heterozygote deficiencies at several loci simul- taneously. Those authors argued the difficulty in separating ho- mozygotes and heterozygotes at all loci, at the same time, "with- high heterozygosity) N heterozygosity i high heterozygosity low heterozygosity high heterozygosity V heterozygosity ; NOV-DEC ; JAN-FE8 MAR-APR ; ill! '!^sawjJi^^*i^wAaMiK■^^R3 '^MifM.'ifS^^tiimSS^ W^. I ■ I n,.a:tfr-ie . i »■ i '^'^ Halden & V. barbin. 1995. lonoluminescence: A new tool tor nuclear Farmer (ed.). The Infrared Spectra of Minerals. Mineralogical Society u /- i c i» n >-i £-> ■^ ^ -^ microprobes in Geology. Scanning Microsc. 9:43-62. Monograph 4. Bartholomew Press. Surrey. ^aremba. C. M.. A. M. Belcher. M. Fritz. Y. Li. S. Mann. P. K. Wilbur. K. M. & A. S. M. Saleuddin. 1983. Shell formation, pp. 235- Hansma. D. E. Morse. J. S. Speck & G. D. Stucky. 1996. Critical 287. In: A. S. M. Saleuddin and K. M. Wilbur (eds.). The Mollusca. transitions in the hiofahncation of abalone shells and flat pearls. Chem. vol. 4. Physiology, part 1. Academic Press Inc.. New York. Mater. 8:679-690. Journal of Shellfish Research. Vol. 15, No. 3. 667-672. 1996. TWO PARASITIC COPEPODS, PSEUDOMYICOLA SPINOSUS AND MODIOLICOLA GRACILIS, ASSOCIATED WITH EDIBLE MUSSELS, MYTILVS GALLOPROVINCIALIS AND MYTILUS CALIFORNIANUS, FROM BAJA CALIFORNIA, NW MEXICO JORGE CACERES-MARTINEZ.' REBECA VASQUEZ-YEOMANS,' AND EDUARDO SUAREZ MORALES^ 'Centra de Inveslii;acidn Cientifica y de Educacion Superior de Ensenada Departamento de Acuicultura Apdo. Postal 2732. 2800 Ensenada, Baja California, Mexico 'El Colegio de la Frontera Sur Apdo. Postal 424 Chetumal. Quiniana Roo, Mexico ABSTRACT Mussel culture and fisheries are two increasing activities in Baja California. NW Mexico. One of the nsks for these activities is the presence of harmful parasites like certain copepod species. This study was carried out to determine the parasitic copepods associated with edible mussels. Mylilus galloproviiicialis Lmk. and Mylilus califoniiainis Conrad, from Baja California. NW Mexico, and to establish certain aspects of their distribution, temporal fluctuation, and damage to their host. The study was carried out at sites of contrasting environmental conditions; exposed rocky shores, protected shores, protected and polluted areas, and culture area. Two species of parasitic copepods were found inhabiting the mantle cavity and gills of mussels; Pseudomyicola spinosiis Raffaele and Monlicelli (Mycolidae) and Modiolicola gracilis Wilson (Clausidiidae). This is the first record of these copepods in Mexican waters. M. gracilis was found in M. galloprovincialis and M. californiamis from all localities studied in numbers from 0 to 5 individuals per host and a maximum prevalence of 26.66% in the first species, and from 0 to 15 specimens per mussel and a maximum prevalence of 70% in the second species. Its presence was relatively constant through the year, with a slight increase in autumn and winter. P spinosus. by contrast, was scarce or absent in M. californiamis and M. galloprovincialis from exposed rocky shore environments. Its number and prevalence were low in the mussel culture area. However, it was very abundant in M. galloprovincialis from the projected and polluted environments, where its numbers ranged from 0 to 59 copepods per mussel and a prevalence of 100%. Rates of infestation in mussels increased in autumn. Macroscopical damages associated with the presence of copepods were not detected, and the histopathologic analysis did not reveal any damage to the tissues of the host. However, there were more parasites in larger mussels, and most parasitized mussels showed a low condition index. P. spinosus could be considered a potential threat to the mussel culture. KEY WORDS: Pseudomvicola spinosKS. Modiolicola gracilis. Mylilus californiamis. Mylilus galloprovincialis. copepods, parasites, pathology INTRODUCTION distribution, prevalence, and effects of parasite copepods on bi- , ,, , valve molluscs arc unknown. The aim of this study was to ascer- The California mussel, Mvtilus calijoniicmus, locally named . , r j ,■ x u j i . i j „, ,, , ,■ w , ,, • .■ 1 '^"1 'he specific identity ot the copepods irom natural and com- "Choro, and the blue mussel, Mv7;/!«ea//oprai'(«ao/M. are used . , , ,. ,.,, , w ,, ; ah ^ , . „ . ^' ,.- K,,,,.. ^, r- mercial stocks ot edible mussels. M. galloprovincialis and M. tor human consumption in Baia Cahtomia, NW Mexico. Ihetirst ,.,. . i j- . i. .• ■ r> • ,^ if • ^ . ' , , . , ,„^^ calitonuaniis: to document their distribution in Baja California, species has been aathered for centuries in the region (Linik 1977. , . ,„ j .u rr . .u j-,- ■ a J;,, „„ . ,. . , ^. . ^, ,. their temporal fluctuation, and their effects on the condition index Tellez 1987) and actually supports a local fishery. The second is ... , , , . .u i et , ,u u ^, ^ '^'^,. , . , / . , , ol the mussels; and to describe any pathologic effects on the host, cultured with submerged longlines and is sold to national and international markets. Mussel culture activities are expected to MATERIALS AND METHODS increase in Baja California. However, there are no studies con- cerning the parasites affecting both edible mussel species in the Seasonal fluctuations of copepods were studied between Janu- area, as well as their potential hazard for commercial production. ary and December 1995 in mussels from Baja California, NW Our previous analysis revealed that some copepods could be found Mexico. During this period, 30 adult M. galloprovincialis were in the mantle and gills of mussels. The poecilostomatoid family collected each month from La Mina del Fraile, an exposed rocky Mycolidae includes 15 species of copepods; most of them are shore (mean total shell length, 57.0 mm; standard deviation [SD], parasitic in marine bivalve molluscs. They have been identified as 5.76); Tres Hermanas. a mussel culture area (mean total shell agents of mass mortality of some commercial species in different length, 64.0 mm; SD, 6.39); and Ensenada Pier, an urban and regions of the world and have been considered to be pests iKor- polluted area (waste water discharge and fuel from ships) (mean ringa 1951, Davey et al. 1978, Ho 19951. In some countries of total shell length, 58.0 mm; SD, 9.02) (Fig. I). Additionally, 30 Hurope and Asia, the marine copepod parasite launa is relatively adult M. califoniiainis were gathered monthly from a fishery area well known (Avdeev 1977. Davey et al. 1978, Paul 1983, Do and in La Mina del Fraile (mean total shell length, 63.0 mm; SD, Kajihara 1986, Theisen 1987, Ho and Kim 1991, Poquet et al. 8.17). In May 1995 and March 1996, other localities were sam- 1994, Robledo et al. 1994). In Mexico, however, the identity, pled to extend the information on the distribution of parasitic cope- 667 668 Caceres-Martinez et al. ■31° 50' 1 16° 45 Todos Santos Bay 1 16° 05' Ensenada Todos Santos Vj ^ — i Islands Pacific Ocean -31° 20' Baja California Baja California Bahia de San Quintin • Mylilus califomianus CZ] Mytilus galloprovincialis Figure 1. Map of the study area showing collection sites and species. I, Inner cliff of Ensenada (polluted area); 2, Ensenada Pier (polluted area); 3, Estero Beach (sandy and estuarine area); 4, Tres Hermans (mussel culture area); 5, La Bufadora (exposed tourist area); 6, El Acuario (exposed nonexploited area); 7, La Mina del Fraile (exposed exploited area); 8, San Quintin (sandy and protected area); 9, San Quintin (exposed exploited area). Baja California. (^Localities where annual study was carried out). pods: M. galloprovincialis from Estero Beach, a sandy and estu- arine locality; M. califomianus from La Bufadora. a tourist, ex- posed rocky shore; M. califomianus from El Acuario, a nonexploited exposed rocky shore; M. galloprovincialis and M. califomianus from the inner cliff of Ensenada. a protected and polluted area; M. califomianus from an exposed rocky shore in Bahfa de San Quintin, an exploited area; and M. galloprovincialis from the inner Bahia de San Quintin. a protected area (Fig. 1). After any fouling organisms were removed, each mussel was sized (total shell length) and weighed (total weight [TW|), after which they were placed in Petri dishes and opened. Intervalvar water and mussel flesh were examined for the presence of cope- pods under a dissecting microscope. Copepods were preserved in a solution of 70% ethanol for identification. Body size, length: width ratios, and length:egg sac length ratios were obtained for all of the identified copepods. Measurements were taken with a mi- crometer eyepiece fitted in a stereoscopic microscope. TW. wet meat weight (MW). and shell weight (SW) of mussels were re- corded to obtain a condition index (CI) where CI = (MW/(TW - SW)) X 100 (Aguirre 1979). Prevalence was estimated as the number of infested mussels/number of mussels examined xIOO. In November 1995, 226 M. galloprovincialis. ranging in size be- tween 20 and 75 mm, were collected from Ensenada Pier for an analysis of the number of parasite copepods and their relation to mussel size (linear regression). The relationship of copepods and the CI of mussels was analyzed in 175 mussels ranging from 60- to 70-mm shell length (linear regression). In both situations, mus- sels were grouped in length classes and the mean number of cope- pods per length class, or CI of mussels, was calculated. For the study of tissue damage, fractions of gills, mantle, and labial palps, in mussels where copepods were found, were re- moved and fixed in Davison's fixative (Shaw and Battle 1957) for 24 h; tissue samples were then embedded in paraffin wax and sectioned at intervals of 5 p-m, and histologic sections were stained with hematoxylin and eosin. Slides were examined under an op- tical microscope at 40 x and 20 x magnifications. RESULTS Identity of Copepods Pseudomyicola spinosus One of the copepod species found in M. galloprovincialis and M. califomianus was identified as Pseudomyicola spinosus (Raf- Parasitic Copepods of Edible Mussels 669 fade and Montieelli 1885). The material studied included 9 adult females, 10 adult males, and 15 female copepodids from the gills and mantle cavity of W. galloprovincialis from Ensenada Pier and from M. californiamts from the inner cliff of Ensenada. All spec- imens were ethanol preserved, and vials are deposited in the United States National Museum, Smithsonian Institution (USNM- 274221). For females, the morphological features of P. spinosiis from M. galloprovincialis and A/, californiamis agree with the diagnosis of P. spinosiis as redescnbed by Ho ( 1980) and Do and Kajihara (1986). Adult female length: width ratios were from 3:4:1 to 4:2:1. length:egg sac length ratios were from 2.5:1 to 2.7:1, and the number of eggs was from 20 to 32. There were two female spec- imens with atypical size, length:width ratios (4.4:1 and 3.2:1), length:egg sac length ratios (3.3:1 and 1.7:1), and number of eggs (22 and 26). The bodies of these specimens were I8'7( longer and 15% broader than the "normar" forms, and their egg sacs were almost 10% longer (see Do et al. 1984 for atypical forms). The male morphology of the studied specmiens agrees with previous descriptions of the species: adult length:width ratios were from 3.2:1 to 3.7:1. As recorded for the females, two aduh male specimens showed variation in their body proportions:length: width ratio, 4.4:1. Atypical males were more slender than "'nor- mal" individuals, being 12% longer and 9% narrower. ModioUcola gracilis The other copepod found in M. californiamis and M. gallopro- vincialis was identified as ModioUcola gracilis (Wilson 1935), which was first reported from Monterey Bay, CA. This is another poecilostomatoid parasitic copepod belonging to the family Clau- sidiiade. Although this is a common copepod in California popu- lations of Mytilus. this is the first record of the species in Mexico. The material studied included four adult females and two adult males from the gills and mantle cavity of M. californiamis and M. galloprovincialis from La Mina del Fraile. All specimens were ethanol preserved, and vials are deposited in the United States National Museum. Smithsonian Institution (USNM-274222). Seasonal Fluctuation and Distribution The prevalences of the copepod species at the study localities are shown in Figure 2. The prevalence of W. gracilis was different among all localities and species, except between M. galloprovin- cialis from Tres Hermanas and M. californianus from La Mina del Fraile (Kruskal-Wallis test, H = 19.7, p < 0.01, followed by all -o o a. OL. •^ •a to 1 I a. o u •s o « at: Mina del Fraile (exposed rocky shore) Mytilus galloprovincialis B Mina del Fraile (exposed rocky shore) Mytilus californianus Tres Hermanas (culture area) Mytilus galloprovincialis Ensenada Pier (polluted area) Mytilus galloprovincialis Figure 2. Prevalence of Af, gracilis (■) and P. spinosus (D), mean number of copepods per sampled mussels (ratio of copepodsrmussel), and CI in M. galloprovincialis from (A) the exposed rocky shore of La Mina del Fraile (exploited area), (C) the mussel culture area of Tres Hermanas, and the (D) polluted area of Ensenada Pier and in M. californianus from (B) the exposed rocky shore of La Mina del Fraile (exploited area), Baja California, Mexico, from January to December 1995, Differences in prevalence and CI were statistically significant (p > 0.01) (see text). 670 Caceres-Martinez et al. pairwise multiple comparison procedures, Dunn's method). The prevalence of P. spinosiis was different among all localities and species, except between M. f>alloprovincialis and M. cctUfornianus from La Mina del Fraile (Kruskal-Wallis test, H = 36.18, p < 0.01, followed by all pairwise multiple comparison procedures, Dunn's method). M . gracilis was currently found in La Mina del Fraile and Tres Hermanas. However, this species was very scarce in Ensenada Pier. In La Mina del Fraile, the prevalence of M. gracilis increased in autumn. From spring to middle summer, it was not observed in M. galloprovincialis. An increase was not detected in the prevalence of M. gracilis during the autumn in M. galloprovincialis from the culture area, but the presence of this copepod continued throughout the year, with the exception of August. By contrast, P. spinostis was occasionally found in mus- sel species from La Mina del Fraile and was frequent in the mussel culture area, but at low prevalence. Interestingly, P. spinosKS was recorded throughout the study period in M. galloprovincialis from the Ensenada Pier, and its prevalence was around 80%. Figure 2 also shows the mean number of copepod species per infested mussel in each locality studied. Variations in the mean number of M. gracilis per infested individual were closely related to the variation in its prevalence throughout the study period. The maximum mean number of these copepods occurred in autumn in La Mina del Fraile, when the maximum numbers of A/, gracilis in M. galloprovincialis and M. californianus were 5 and 15, respec- tively. In Tres Hermanas, there were tluctuations in the mean number of copepods per host that were closely related to fluctua- tions in prevalence. The maximum number of copepods per M. galloprovincialis was three. The range and mean number of P . spinosus per infested mussel in the Ensenada Pier were higher from summer to autumn than the rest of the year. The maximum number of copepods per mussel was 59 in October 1995. Oviger- ous females of M. gracilis were recorded in all samples from La Mina del Fraile and from Tres Hermanas. Ovigerous females of P. spinosus were found throughout the study period in the Ensenada Pier, where copepodid stages were also recorded in September and October 1995. Both copepod species, M. gracilis and P. spinosus. were found together in the same host. There were differences in the CI of mussels from localities and species throughout the study period (analysis of variance, F = 9.51. p < 0.001, followed by Student-Newman-Keuls mean comparisons method). There were no significant differences between the CI of M. galloprovincialis and M. californianus from La Mina del Fraile and between the CI of W. galloprovincialis from La Mina del Fraile and Tres Herma- nas, but significant differences were recorded between M. gallo- provincialis from Tres Hermanas and M, californianus from La Mina del Fraile, where the prevalence and mean number of cope- pods per mussel were similar. The CI of M. galloprovincialis from the Ensenada Pier area was less than those of mussels from the other localities, and this was statistically significant (Fig. 2). Table 1 shows the results of the presence of copepod species in the study area. M. gracilis is a common copepod associated with both M. californianus and M. galloprovincialis. It was found in M. californianus from La Mina del Fraile, El Acuario, Bahia de Todos Santos, and Bahi'a de San Quintin. In M. galloprovincialis this copepod was found in all of the localities studied (see also Fig. 2). In both mussel species, the prevalence and range of M. gracilis were low. Contrary to this, P. spinosus was very scarce on ex- posed rocky shores but it was very abundant in number and prev- alence in protected and polluted environments like the Ensenada Pier. There was a significant positive correlation between the size of A/. ^?«//oprov/nf/a/(.j and the presence of copepods (y = —6.98 -I- 0.28x,R- = 0.81, F = 37,p<0.1). Mussels from 45- to 75-mm shell length had the highest number of copepods. The lowest CI of M. galloprovincialis was recorded in the Ensenada Pier, the most copepod-infested area. The relation between low CI in mussels with the highest number of copepod parasites was corroborated by the correlation between the number of copepods and the CI of their host in this locality (y = 34.13 - 0.51x, R- = 0.86, F = 18.6, p < 0. 1 ). Neither macroscopic analysis nor histologic analysis of mussel gills, labial palps, or mantle revealed any damage. DISCUSSION This is the first record of P. spinosus and M. gracilis in Mex- ican waters. M. gracilis has been found in M. galloprovincialis and Mylilus ediilis L. from European and Japan waters and in M. ediilis from Monterey Bay, CA (Wilson 1935, Ho 1980, Poquet et al. 1994). In accordance with Poquet et al. (1994), this species does not show any morphological adaptations to parasitism; how- ever, it presents some characteristics in its cuticule (external mi- crovilli-likc projections) that suggest a mechanical function in ad- hesion rather than nutritional absorptive function. Histologic stud- ies, however, have failed to show any histopathologic lesions. Poquet et al. ( 1994) pointed out that the determination of specific functions of the integumental glands and the characterization of the chemical composition of the various cuticular layers could lead to the identification of possible long-term alterations in host tis- sues. The recorded prevalences of M. gracilis in this study were higher than those recorded in M. galloprovincialis and M. edulis from the Ebro Delta River (E. Spain) (Poquet et al. 1994). M. gracilis was not recorded in August, when the highest tem- peratures prevailed. The slight increase in the prevalence of this parasite recorded from September to December in La Mina del Fraile suggests that the reduced temperatures that occur during these months may be favorable for this parasite. Permanently sub- merged mussel culture conditions seem to be favorable for the regular prevalence of M. gracilis. P. spinosus has been associated with 54 species of bivalve molluscs in temperate and tropical waters (Ho 1995, Humes 1968, Ho and Kim 1991). Measurements of the Baja California speci- mens revealed variations among adult and preadult individuals. The size of normal female specimens falls within the range known for P. spinosus: however, their bodies are relatively longer and more slender than specimens from Japan. The egg sac proportions of Japanese and Baja California specimens showed only slight differences (Do et al. 1984). The atypical female is noticeably longer than normal individuals and can be distinguished by its larger size and relatively longer egg sacs than those of the "nor- mal" females. In the Japanese material (Do et al. 1984), the proportion of body length:egg sac length is about 3:1, whereas in the atypical form, the value is only 1 .75: 1 . Atypical males of Baja California were comparable to Japanese atypical males (Do et al. 1984). In both cases, the body is more slender than in the normal individuals, and when comparing the length:width ratio of both atypical forms, Baja California specimens appear to be even more slender (4.4:1) than the Japanese specimens (3.9:1). Do et al. ( 1984) suggested that these slender body forms appear to be char- acteristic of active swimmers. However, we did not observe P. spinosus swimming: it remained crawling through the mantle and gills. Parasitic Copepods of Edible Mussels 671 TABLE 1. Copepods infecting M. californianus and M. galloprovincialis from sampling localities in Baja California, Mexico, in May 1995* and March 1996#." Mussel Range of Mean Length Cope pod Mean per Locality Species n (mm) SD Species Infected Mussel *La Mina del Fraile. exposed rocky shore M^ californianus 30 60,47 7.71 0-1 1 (exploited) M. gracilis *La Mina del Fraile. exposed rocky shore M. galloprovincialis 30 42.31 4.90 0 0 (exploited) *E1 Acuario. exposed rocky shore (not M. californianus 29 66.91 10.51 0-3 1.5 exploited) M. gracilis *La Bufadora, exposed rocky shore M. californianus 30 51.59 7.59 0-1 I (touristic) M. gracilis *Tres Hermanas (culture) M. galloprovincialis 30 71.27 9.21 0-2 M. gracilis 1.2 *Estero Beach (sandy and estuanne) M. galloprovincialis 30 56.69 8,38 0-4 M. gracilis 2 *Ensenada Pier (polluted) M. galloprovincialis 30 61.75 6.13 0-15 P. spinosus 5.6 #lnner Cliff of Ensenada (polluted) M. galloprovincialis 30 58.23 5.46 0-20 P. spinosus 0-4 M. gracilis 5.8 1 7 #Inner Cliff of Ensenada (polluted) M. californianus 30 64.32 7.60 0-11 P spinosus 0-2 M. gracilis 2.8 1.3 #San Quintin, exposed rocky shore M. californianus 30 60.12 7.23 0-3 1.7 (exploited) M. gracilis #San Quintin, sandy and protected M. galloprovincialis 15 57.89 8,12 0-2 M. gracilis 1.5 ■" Mussel mean length (n = number of mussels examined). SD. range of copepod species per sample, and mean number of copepod species per infested mussel are shown. As we found. Do and Kajihara ( 1986) also detected ovigerous females and adult males of P. spinosus throughout the year in Mytilus ediilis galloprovincialis Lmk. In accordance with those authors, we found a relatively constant prevalence of this parasite through the year. Do and Kajihara (1986) estimated that at least five to six generations of P. spinosus can coexist in a year. They reported a higher prevalence and mean number of parasites per mussel (85% and 2.4, respectively) than in this study. The exact pathologic effects attributable to several parasitic copepods have remained largely uncertain (Do and Kajihara 1986). In the case of a notorious, intestinal copepod, Mylilicola inleslinalis Steuer, Cole and Savage ( 1951), Sparks (1962). Dare ( 1981 ), and Paul ( 1983) have recorded the adverse effects of this parasite on the condition of mussels, M. edulis: the European oyster, Ostrea edulis L; and the Japanese oyster, Crassoslrea gi- gas Thunberg. In the case of P. spinosus, only Dinamani and Gordon (1974) reported the mechanical and pathologic effects in the gut epithelium of the rock oyster. Crassoslrea glomerata Gould, but no definite evidence of damage to the gills or labial palps was found in any oysters examined (Do and Kajihara 1986). Although we did not find any macroscopical or microscopical damage in the gills, labial palps, and mantle of the host, the CI of mussels from the Ensenada Pier, where copepods were very abun- dant, was the lowest of all mussels studied throughout the year. It is possible that damage to the host cannot be detected by macro- scopical and microscopical studies of the gills, labial palps, and mantle, but that it may be assessed at the physiologic level or by studying the gut of the host, where this copepod may be found (Korringa and Lamberg 1951). The reduction in the CI in heavily infected mussels for P . spinosus represents a potential thread to mussel culture if the P. spinosus population increases under cul- ture conditions. The transfer of mussels from infested areas to noninfested culture areas must be avoided. The occurrence of both P. spinosus and M. gracilis in the same host (M. californianus or M. galloprovincialis) indicates that both copepod species may coexist. The presence of the different cope- pod species in the same host has been observed previously in populations of M. edulis in Newport Bay (Ho pers. comm.). The number of copepods seems to be related to larger organ- isms. Similar observations on this relation have been recorded by Costanzo and Calafiore (1987) in Modiolicola insignis Aurivillius; they pointed out that smaller mussels (under 33-mm shell length) were likely to escape infestation. Further studies are being earned out in order to determine the reasons for the differential prevalence in the host mussels and under environmental conditions and to determine whether P. spinosus and M. gracilis are found in the gut of M. californianus and M. galloprovincialis and are producing any lesions. ACKNOWLEDGMENTS We thank Dr. J. A. F. Robledo and the anonymous reviewers for critical advice. Thanks to K. Castaheda, L. Monrov, and P. 672 Caceres-Martinez et al. Macias for their help in sample processing and to S. Guevara from the mussel culture company for allowing and facilitating sampling in their installations. Also, we thank L. Figueroa for typing the data series and N. Flores and F. Valenzuela for their help in field sampling. We thank Dr. Ho for confinmng the identity of the copepod species and for his comments. Thanks to L. Morales from the CICESE library for providing us with the bibliography. This work was supported by the inner project of CICESE # 623106. LITERATURE CITED Aguirre, M. P. 1979. Biologi'a del mejillon (Mviilns edutis) de cullivo de la Ria de Vigo. Bol. hut Esp. Oceanogr. 5:107-160. Avdeev. G. V. 1977. Copepods (Cyclopoida) parasitic on bivalve mol- luscs in Pos'et bay in the sea of Japan. Soviet J- Mar. Biol. 3:108-1 17. Cole, H. A. & R. E. Savage. 1951. The effect of the parasitic copepod, Mylilicola intestinalis (Steuer) upon the condition of mussels. Parasi- tology 4\:\56-l6\ . Costanzo, G. & N. Calafiore. 1987. Seasonal fluctuation of Modiolicola insignis Aurvillius, 1882 (Copepoda: Poecilostomatoida: Sabelliphi- lidae), associated with Mytilus galloprovincialis in Lake Faro (Messina). J Crustacean Biol. 7:77-86. Dare, P. J. 1981. The susceptibility of seed oysters of Ostrea edulis L. and Crassostrea gigas Thunberg to natural infestation by the copepod Mylilicola intestinalis STEUER. Aquaculture 26:201-211. Davey, J. T., J. M. Gee & S. L. Moore. 1978. Population dynamics of Mylilicola intestinalis in Mytilus edulis in South West England. Mar. Biol. 45:319-327. Dinamani, P. & D. B. Gordon 1974 On the habits and nature of asso- ciation of the copepod Pseudomyicola spinosus with the rock oyster Crassostrea glomerata in New Zealand. J Invertebr. Pathol. 24:305- 310. Do, T. T. & T. Kajihara. 1986. Studies on parasitic copepod fauna and biology of Pseudomyicola spinosus, associated with blue mussel, Mytilus edulis galloprovincialis in Japan. Bull. Ocean Res. Inst. Univ. Tokyo 23:\-63. Do, T. T., T. Kajihara & J.-S. Ho. 1984 The life histor>' of Pseudomv- icola spinosus (Rafaele and Monlicelli, 1985) from the blue mussel, Mytilus edulis galloprovindialis Lamarck in Tokyo Bay, Japan, with notes on the production of atypical male. Bull. Ocean Res Inst. Univ. Tokyo 17:1-65. Ho, J. S. 1980. Origin and dispersal of Mytdus edulis in Japan deduced from its present status of copepod parasitism. Publ. Seto Mar. Biol. Lab. 25:293-313. Ho, J. S. 1995. Mycolid copepods and mass mortality of culture hard clams (Merethrix). Sixth International Workshop of the Tropical Ma- nne Mollusc Programme, June 12-20, 1995. Annamalai University, Parangipettai , India (abstract). Ho, J. S. & I. H. Kim, 1991. Copepod parasites of commercial bivalves in Korea. II. Copepods from cultured bivalves Bull Korean Fish. Soc. 24:369-396. Humes, AG. 1968. The cyclopoid copepod Pseudomyicola spinosus (Raffaele and Monlicelli) from marine pelecypods, chiefly in Bermuda and the West Indies. Beaufortia 14:203-226. Komnga, P. 1951 . De aanval van de parasiet Mylilicola intestiiuilis op de Zeeuwse mosselcultuur. Visserijnieuws 7:1-7. Korringa, P. & L. Lambert. 1951 . Quelques observations sur la frequence de Mylilicola intestinalis Steuer (Copepoda parasitica) dans les moules du littoral mediterraneen francais avec une note sur la presence de Pseudomyicola spinosus (Raff. & Mont.) (Copepoda parasitica). Revue Trav. (Sclent. Tech.) Pech. Maril. 17:15-29. Linik, T. W. 1977. La Joya natural radiocarbon measurements VII Ra- diocarbon 19:19^8. Paul, J. D. 1983. The incidence and effects of Mylilicola intestinalis in Mytilus edulis from the Rias of Galicia, North West Spain. Aquacul- ture 3l:\-lO. Poquet, M., E. Ribes, M. G. Bozzo & M. Durfort. 1994. UltrastrucUire and Cytochemestry of the integument of Modiolicola gracilis, parasitic copepod in mussel gills {Mytilus galloprovincialis and Mytilus edulis. J Morphol. 221:87-99. Rafaelle, F. & F. S. Monlicelli. 1985, Descrizione di un nuovo Licho- molgus parasita del Mytilus galloprovincialis Lmk. Atti R. Accad. /.mrei 4:302-307. Robledo, J. A. P., M. M. Santarem & A. Figueras. 1994. Parasite loads of rafted blue mussels {Mytilus galloprovincialis) in Spain with special reference to the copepod, Mylilicola intestinalis. Aquaculture 127: 287-302, Sparks, A. K. 1962. Metaplasia of the gut of the oyster Crassostrea gigas (Thunberg) caused by infection with the copepod Mylilicola orientalis Mori. J. Insect Palhol. 4:57-62. Suh, H.-L. & S.-D. Choi. 1990. Two copepods (Crustacea) parasitic on the blue mussel, Mytilus galloprovincialis. from the Yongsan river estuary m Korea. Bull. Korean Fish. Soc. 23:137-140. Shaw, B, L. & H. L. Battle. 1957. The gross microscopic anatomy of the digestive tract of the oyster Crassostrea virginica (Gmelin). Can. J. Zool. 35:325-346. Tellez, D. A. 1987 Los concheros de Baja California y sus perspectivas de investigacion. Int. Inv. Soc. Eslud. Fronterizos 5:111-116. Theisen, B. F. 1987. Mylilicola intestinalis Steuer and the condition of its host Mytilus edulis L, Ophelia Tl-.ll-id. Wilson, C, B, 1935, Parasitic copepods from the Pacific Coast. Am. Mid. Nat. W.ll 6-191 . Journal of Shellfish Research. Vol 15. No. 3. 673-680. 1996. MICROGEOGRAPHIC VARIABILITY IN FEEDING, ABSORPTION, AND CONDITION OF MUSSELS (MYTILUS GALLOPROVINCIALIS LMK.): A TRANSPLANT EXPERIMENT J. I. P. IGLESIAS,' A. PEREZ CAMACHO,^ E. NAVARRO,' U. LABARTA,' R. BEIRAS,^ A. J. S. HAWKINS,^ AND J. WIDDOWS^ ^Departamento de Biologia Animal v Genetica Facultad de Ciencias Univeisidad del Pais Vasco I Euskal Herriko Unibertsitatea Apdo 644 E-48080 Bilbao, Spain 'Institute Espanol de Oceanografia Muelle de Animas s.n. Apdo. 130 E-15001 La Coruna. Spain ^Consello Superior de Investigacions Cienti'cas Instituto de Investigacions Marinas Eduardo Cabello 6 E-36208 Vigo. Spain '^Plymouth Marine Laboratory Prospect Place Plymouth PL I SDH. United Kingdom ABSTRACT Mussels, Mylilus galloprorincialis Lmk.. were reciprocally transplanted between three cultivation rafts in the Ria de Arousa (Galicia. northwest Spainl. After an 8-wk period, rafts were visited and individual measurements of clearance rate, absorption efficiency, and condition index were performed in the field under ambient conditions of temperature, salinity, and food availability. Clearance rate standardized to a common shell length was not significantly affected by the location of mussels in the Ria but varied according to their origin. Absorption efficiency was mainly affected by raft position, reflecting the spatial variability in the quality of available food. Origin-related differences in absorption efficiency showed the same trend as recorded for clearance rates. No gradient of environmental factors has been recorded in the Ria that might account for differences in physiologic parameters persistent after transplantation (i.e.. origin effects). Associated evidence suggests that observed differences may have resulted from spatial variability in the degree of parasitic infestation. Condition index was dependent on both raft position and mussel origin and closely reflected the described vanability in physiological parameters. Remaining variability in condition index can be attributed to different conditions before transplantation took place. KEY WORDS: Feeding, absorption, condition, transplant. Mylilus INTRODUCTION clearance rate (CR) and AE measured in the laboratory under a standardized feeding regimen (Hawkins el al. 1985) or by differ- The rate of feeding and absorption efficiency (AE) constitute ences in the CR of individuals from populations inhabiting media the physiologic parameters controlling energy intake in most ani- characterized by very different seston concentrations and compo- mal groups. Marine bivalves exhibit great variability in these pa- sitions (Theissen 1977, Bayne et al. 1984, Bayne et al. 1987). rameters in response to different types of factors, both environ- Another important factor affecting growth potential has been re- mental and endogenous. Although rates of feeding are primarily cently reviewed by Newell and Barber (1988) and refers to the dependent on the availability of food resources, relationships be- negative effect of diseases and parasites on parameters controlling tween ingestion rate and particle concentration are mediated by energy acquisition. Although studies dealing with this aspect are clearance or pumping rate, and this has been found to depend on scarce, preliminary results have begun to provide the physiological a variety of environmental factors (see review by Bayne and New- basis for understanding the deleterious effect that diseases and ell 1983, Hawkins and Bayne 1992). Absorption efficiency is parasitic infection exert on growth rate and reproductive potential, mainly affected both by the concentration (negatively) and quality In addition, genetic factors have been shown to influence physi- (positively) of suspended food matter (Bayne and Newell 1983, ological variability between individuals of the same population Bayne effl/. 1987, Bricelj and Malouf 1984, Navarro e/ a/. 1991, (Koehn and Shumway 1982, Hawkins et al. 1986). However, in Navarro et al. 1994). transplant experiments designed to simultaneously test the effects Variability in parameters of energy gain has been interpreted in of genetic and environmental factors on growth rates, local envi- some instances as an adaptation to the specific feeding conditions ronmental conditions affected the major proportion of recorded characteristic of the environment where mussels live. Such adap- differences (Dickie et al. 1984, Mallet and Carver 1989), but see tations have been evidenced, for example, by seasonal cycles of Peterson and Beal (1989) for a somewhat different result. 673 674 Iglesias et al. When trying to quantify the extent to which observed differ- ences in the physiological components of energy balance in mus- sels from different locations are dependent on environmental fac- tors, reciprocal transplantations appear to be the best-suited pro- cedure. Widdows et al. (1984) compared the seasonal cycle in energetic parameters of native and reciprocally transplanted mus- sels, Mytilus edulis, from two populations. After 2 mo, the clear- ance rate of transplanted mussels differed by only 20% from those in native mussels, but the remaining differences persisted even after another 3 mo. Authors suggested that either full acclimation needed longer periods of time, or that some genotypic component was responsible for the residual difference. In contrast with these results. Okumus and Stirling (1994) have recently observed that physiological differences between mussels cultivated at two Scot- tish sea lochs disappeared after a 4.5-mo period of reciprocal transplantation. In a previous study (Navarro et al. 1991), we reported differ- ences in parameters of feeding and absorption for mussels {.Mytilus galloproviiuialis) growing on different sites in the Ri'a de Arosa. This study describes transplant experiments that were undertaken to discriminate between alternative causes of these differences, associated either with the present environment or with different origins. Mussels cultivated in rafts at three sites were reciprocally transplanted, and after 8 wk, CR and AE were measured in the field under ambient conditions of temperature, salinity, and food availability. The degree of persistence of physiological differences after transplantation was assumed to provide some insight on the nature of those differences. Condition index was also determined to provide an independent measure of physiological status. MATERIAL AND METHODS Animals, Sites and Transplantations Experiments described in this study were performed in Novem- ber 1989. According to previous information on the physiological behavior of cultured mussels (M. gcilloprovincialis Lmk.). three rafts located in the Ria de Arosa (Galicia. northwest Spain) were chosen for this study (see Fig. I ). Raft A was moored at the head of a huge grouping (141 rafts) located in an area of maximum oceanic influence within the Ria. Raft B belonged to a small grouping ( 16 rafts) close to the Isla de Arosa, a zone that has been identified by Otto (1975) as an upwelling area characterized by high values of primary production. Raft C was located in a group- ing (40 rafts) positioned within the inner part of the estuary, where oceanic influence is minimal. Eight weeks before physiological measurements, 150 individ- ual mussels were randomly sampled from a rope located in the central part of the front of each raft and were divided into three groups that were each placed within a separate net bag. One bag remained immersed in the raft of ongin, and the other two were transplanted, one to each of the other two rafts. After this proce- dure, 50 mussels of each origin (Rafts A, B, and C) remained immersed under the same environmental conditions in each of the three rafts (A, B, and C) during the transplant period. Bags were located in the central part of the front of each raft and at a I -m depth. At the end of the transplant period, rafts were visited on consecutive days and measurements were performed under ambi- ent conditions. Temperature and salinity remained similar between rafts at the time when physiological determinations were made (mean temperature = 15.37 ± 0.52°C; mean salinity = 35.2 ± 0.43%) (± I SD). Figure 1 . Detailed map of the Ria de Arousa showing the sites where rafts were located. Squares denote raft groupings. Procedures Twelve mussels were taken from each bag and arranged in individual trays within a feeding tank. Water collected from the same depth where mussels had previously been hung was pumped to the feeding tanks (one per origin) through multiple inflow lines. This arrangement was designed to avoid gradients of seston con- centrations into the tanks. Flow rate was kept above 2.5 L/min because at this rate, differences in seston concentration between tanks caused by variable filtering rates were calculated to be neg- ligible. After 30 min of acclimation to flow-through conditions, dupli- cate 2-L seawater samples were collected at hourly intervals for 4 h. Two hours after the first water sample was taken, the tanks were cleaned and the feces produced by each individual mussel were completely collected on three occasions over the following 4 h. Measurements Seawater samples and aliquots of known volumes from each fecal sample were filtered onto preashed (450°C for 4 h) and weighed GFC filters and rinsed with isotonic ammonium formate. Total dry matter was established as the weight increment deter- mined after drying the filters to constant weight at 1 I0°C. Organic matter corresponded to the weight loss after ignition at 450°C in a muffle furnace. After this procedure, total particulate matter (TPM. mg/L), particulate organic matter (POM. mg/L). and par- ticulate inorganic matter (PIM. mg/L) were determined within the water passing through each feeding tank. The egestion rates of inorganic matter (mg/h) were determined as means of triplicate fecal samples collected from each individual Feeding and Absorption in Transplanted Mussels 675 mussel and were assumed to represent inorganic ingestion rate (i.e.. that there was no absorption of ash in the digestive tract). CR (L/h) were then estimated indirectly, with PIM concentration (mg/L of seawater) as the reference for available inorganic matter. The validity of this procedure for estimating CR has been recently tested in both cockles (Urrutia et al. 1996) and mussels (A. J. S. Hawkins unpublished results) and has been used for measuring CR in mussels under ambient conditions of food availability (Hawkins et al. 1996). AE was measured by the ratio method of Conover (1966), on the basis of organic fractions of food and feces. The same three replicate samples used for measuring egestion rates were used to deteiTnine the mean absorption efficiencies for each individual. After measurements were completed, the soft tissues of each mussel were excised, dried at 85°C, and weighed to the nearest mg (DTW. in grams). After ignition to constant weight at 450°C in a muffle furnace, ash-free dry tissue weight (AFDTW, in grams) was also determined. The length (L) of each mussel shell was measured to the nearest millimeter, and dry shell weight (DSW, in grams) was measured after drying in an oven to constant weight. Condition index (CI) was then calculated for each mussel as: CI 100 X AFDTW/DSW. Size Standardization and Statistical Analysis In general, physiological rates need standardization to elimi- nate size-dependent variability. The most commonly used refer- ence for size is soft body mass; however, the weight standardiza- tion of CR may be somehow arbitrary because this rate is consid- ered to be dependent on filtration (gill) area (Hughes 1969). To obviate problems derived from possible variations of gill area per unit body weight in mussels from different origin, the measure- ments of CR presented here were standardized to both body mass and shell length. Mass-specific ( 1 g) CR was obtained according to the formula: CR, = CR, X (l/W,)", where CR, is the CR of the standard-sized mussel, CR, is the uncorrected CR, 1 is the standard mass ( = 1 g), W, is the mass of the experimental mussel, and b is the power that scales CR with the body mass. In this work, we have used a value b = 0.53, estimated by Perez and Gonzalez ( 1984), for M. galloprnyiiutalis from the Ria de Arosa. For shell length standardization we used the formula: CR, = CR, X (80/L,)^ where CR, and CR, are as in the former expression, 80 is the standard length ( =80 mm), L, is the length of the experimental mussel, and b ( = 1.85) is the power that scales the CR with the shell length for mussels from Arosa (Perez and Gonzalez 1984). Standard analysis of variance (ANOVA) procedures (Zar 1984) were applied to test the simultaneous effect of raft location (Raft) and raft of origin (Origin) on the parameters measured in this work: CR, AE, and CI. RESULTS Seston Characteristics Parameters describing the characteristics of seston at the three rafts are presented in Table I . Recorded TPM values are well below those found for the same rafts in the previous spring: from 1 .065 mg/L in raft 6 (raft A in this work) to 2. 180 mg/L in raft 4 (raft C) (Navarro et al. I99I). Because no data on seasonal changes in seston concentration are available, we cannot conclude whether these differences are due to a seasonal cycle or whether they are merely the consequence of short-term variability in this parameter. However, the organic contents of the seston were very similar to values previously obtained on different occasions for seven rafts within comparable areas of the same estuary (Cabanas et al. 1979, Navarro et al. I99I), and which reflects a remarkable spatial and temporal constancy of this parameter for the Ria de Arosa. Although no direct measurements of particulate volumes were performed on seston samples taken in this work, approximate es- timations of this parameter were obtained by transforming re- corded TPM (in milligrams per liter) to volume (VOL, in cubic millimeters per liter) equivalents using the equation: VOL = 0.4006 + 0.5005 x TPM (r" p < 0.001), 0.832, n 12, which was fitted to previous simultaneous determinations of VOL and TPM (Navarro et al. 1991 ). Food quality was then calculated as Q = milligrams of POM per cubic millimeter, mean values (X ± SD) for which were as follows: Raft A, 0.543 (±0.066): Raft B, 0.484 (±0.044); and Raft C, 0.361 (±0.056). CR Mass-specific CR are presented in Figure 2A (X ± 95% con- fidence interval), and results of ANOVA are in Table 2. Both Raft and Origin exert significant effects on this parameter. However, a comparison of the F values suggests that, among the factors con- sidered here, the main influence was Origin. Maximum and min- imum CR corresponded to mussels from origins B and C, resfwc- tively. When standardization is performed for a common length (Fig. 2B), the trend observed for different Origins remains as described for mass-specific CR. Nevetheless, this factor only explains 37% of the recorded variance in CR (Table 3). The effect of Raft, TABLE 1. Seston characteristics recorded in the course of physiological measurements.^ RAFT TPM (mg/L) POM (mg/L) PIM (mg/L) / A B C 0.702 ± 0.379 0.871 ± 0.076 0.532 ± 0.041 0.408 ± 0.255 0.405 ± 0.054 0.241 ± 0.031 0.294 ± 0.132 0.466 ± 0.052 0.291 ± 0.022 0.568 ± 0.070 0.465 ± 0.042 0.451 ± 0.070 " f, organic content (decimal fraction). Data are means ± SD. 676 Iglesias et al. A RAFT ABC B s 2^ I 1 _] 0 6 _ 5 UJ ABC ABC ABC 1 I- 0 RAFT ABC ABC ABC ABC ORIGIN ORIGIN Figure 2. CR of mussels transplanted from origins A, B, and C mea- sured at rafts A, B, and C. (A) Mass-specific (1 g) CR. (B) Length- specific (80 mm) CR. Vertical lines represent 95% confidence interval. however, becomes nonsignificant, which reflects the higher mass- specific CR of mussels with lower length-specific mass (refer to Discussion). It is worth noting that, in both ANOVA. interaction terms were found not to be statistically significant. Thus, although effect of Raft position in the Ria appears doubtful, the effect of Origin is a significant phenomenon, revealing that either genetic differences or the previous history of mussels elicites fundamental physiological effects. AE Mean AE values for each transplant condition and correspond- ing ANOVA are shown in Figure 3 and Table 4, respectively. Both Raft and Origin exert significant effects on AE. In this case. Raft is responsible for the higher proportion of recorded variance (59%), reflecting the fact that AE is highly affected by the envi- ronmental conditions prevailing in the area where rafts are located. Nevertheless, the effects of Origin were also highly significant. although this factor only explains 99c of observed variance. Even more remarkably, origin-dependent differences in AE exhibit the same trend as described for CR data: the maximum AE corre- sponding to mussels from origin B and the minimum AE corre- sponding to those from origin C. As for CR. the mteraction term was found to be not significant. In Figure 4. we have plotted mean AE values against the seston qualities estimated for each raft from TPM recorded at the time of physiological determinations. Similar measurements obtained in the previous spring for the same three rafts are included for com- parison, together with an exponential curve that had previously been fitted to data from mussels of Arousa (Navarro et al. 1991 ). TABLE 2. Summary of ANOVA for testing significance of differences among mass-specific CR. Factor d.f. SS MS F P Raft 2 6.996 3.498 4.003 0.021 Origin 2 32.090 16.045 18.359 0.001 Interaction 4 3.814 0.954 1.091 0.365 Error 99 86.520 0.874 TABLE 3. Summary of ANOVA for testing significance of differences among length-specific CR. Factor d.f. SS MS F P Raft t 1.370 0.685 0.482 0.617 Origin -} 87.057 43.529 30.656 0.001 Interaction 4 4.479 1.120 0.789 0.535 Error 99 140.573 1.420 The AE values presented here and the included curve follow a similar trend, suggesting that raft-dependent variability in this pa- rameter results from existing differences in the quality of sus- pended particulate matter available to the mussels from each raft. On the other hand, once the effect of variability in food quality is removed, differences in AE associated with Origin alone remain of similar magnitude to those recorded in the previous spring. Condition Index Large differences in CI appear associated with both Raft and Origin (Fig. .5 and Table 5). the interaction term being nonsignif- icant as in previous physiological determinations. The effect of Raft accounts for most variability in CI (27%), revealing a clear gradient between the inner (raft C, CI = 12.75) and outer (raft A, CI = 18.53) zones of the Ria. Although not as great as the effect of Raft, a highly significant influence is also exerted by Origin (Table 5). which explains 20% of recorded variability in CI. DISCUSSION Effect of Raft SS: sum of squares, MS: mean squares, F: F value The effects of Raft on CR were not apparent when standardized for a common shell length, but were significant when standardized RAFT 2 w I— ( u w z o Cl, PC O < -._ ABC ABC ABC ORIGIN Figure 3. AE of mussels transplanted from origins A, B, and C mea- sured at rafts A, B, and C. Vertical lines represent 95% confidence intervals. A B C 0.8 ■ rtf 0.6 - rtf -1 0.4 ■ fi 1 0.2 - Feeding and Absorption in Transplanted Mussels 677 TABLE 4. Summary of ANOVA for testing signiricance of differences among AE. Factor d.f. SS MS Raft Origin Inlcraclion Error 2 2 4 99 3.170 0.499 0.102 1.598 1.585 0.249 0.025 0.016 98.211 15.452 1.578 0.001 0.001 O.lSb to soft-body mass. We suggest that the apparent effects of Raft on mass-specific CR do not represent true differences in the feeding behavior of mussels in different areas of the Ria de Arousa. As bivalves grow bigger, an increasing proportion of produced biomass is allocated into energy stores and reproductive products (this is evidenced by increasing reproductive effort with size or age; Bayne et al. 1983, Peterson 1983, Peterson 1986, Thompson 1984. Iglesias and Navarro 1991). Thus, large variations in total body mass may occur between individuals with structural organs (shell, gills, etc.) of similar size. This appears to be true in this study: the individuals presented here ranged from 68 to 89 mm in shell length (mean. 78 mm), which is close to the maximum size attainable in the Galician Ri'as (90-1 10 mm). Further, during the autumn, a season charactenzed by comparatively low rates of shell growth (Aguirre 1979, Perez and Roman 1979), the mussels used in this study showed dry masses that varied from 698 to as much 0.25 S 0.20 O P 0.10 5 § 0.05 u 0.00 A RAFT B rtl ABC ABC ABC ORIGIN Figure 5. CI of mussels transplanted from origins A, B, and C and measured at rafts A, B, and C. Vertical lines represent 95% confi- dence interval. >- U z y PL, o k; o pa < 0.8 0.6 0.4 0.2 0.0 0.2 0.4 0.6 0.8 FOOD QUALITY (mg. POM mm- 3) Figure 4. AE as a function of food quality (q) recorded al sites of rafts A (q = 0.543). B (q = 0.484), and C (q = 0.361). Points are mean values ±95% confidence interval. Circles, Origin A; squares. Origin B; triangles, Origin C. Solid symbols, data obtained in these experi- ments; hollow symbols, AE recorded in the previous spring (April 1989) for mussels from sites A, B, and C (Navarro et al. 1991). The equation of the curve: AE = 0.807 (1 - e ') (q, food as 4,031 mg (mean. 2,093) and that correlated poorly with shell length (r^ = 0.318 for the log-log relationship). Under these cir- cumstances, of such variable growth of soft tissues, standardiza- tion to a common dry-flesh weight may result in misleading esti- mations of CR. so that the differences observed among rafts would be spurious. In particular, significant higher mass-specific CR in mussels transplanted to raft C probably resulted from their lower dry weights for similar shell sizes. AE was influenced by raft position. As documented in previous studies with species of Mytilus. AE changes in response to short- term variations in the quality of suspended matter, expressed as POM availability per unit particulate volume, both in the labora- tory (Bayne et al. 1987) and in the field ( Navarro etal. 1991). The curve shown in Figure 4 illustrates this dependence for mussels from the Rfa de Arousa. Points included in the figure that represent mean AE for each transplant condition agree reasonably well with the curve. Therefore, we can conclude that differences associated with Raft are the consequence of different food qualities occurring at each raft when experiments were performed. Raft-dependent variability in CI was highly significant. As dis- TABLE S. Summary of ANOVA for testing significance of differences among CI. quality) has been taken from the same previous work (Navarro el al. 1991). Factor d.f. SS MS F P Raft 2 600.862 304.431 26.319 0.001 Ongin 2 462.458 231.229 19.991 0.001 Interaction 4 42.529 10.632 0.919 0.456 Error 99 1145.118 11.567 678 Iglesias et al. cussed above, a major proportion of energy retained by the spec- imens used in this study is very likely devoted to storage reserves or reproductive products. Therefore, recorded differences in CI may have resulted from different growth rates achieved by mussels in each zone of the Ria. Certainly, a close corespondence between growth rate and CI has been previously reported for mussels grown in cultured plots (Smaal and Van Stralen 1990). Because CR per unit shell length remained independent from Raft, maximum growth rates (assumed to be represented by CI) corresponding to Raft A and minimum to Raft C must have resulted from increasing food availability and/or quality with geographic position from in- ner to outer areas of the Ria. This interpretation is supported by associated changes in AE. suggesting that differences in food qual- ity, through its effect on AE. can be considered as the main factor that was responsible for recorded variability in CI. Effect of Origin CR is highly affected by Origin, irrespective of the standard- ization procedure used, and differences associated with this factor present the same trend of variation imposed by Raft: higher CR values for ongin B and lower for origin C. Although AE is mainly dependent on raft position. Origin also exerted a significant influ- ence, mean values of AE showing the same trend as that described for CR. Spatial variability of temperature or salinity is small in Arousa (Landin 1987). Nutritional conditions known to modify physio- logical traits of mussels (Bayne et al. 1984. Bayne et al. 1987), especially food quality, may be more variable among rafts. How- ever, mussels adapt to new nutritional conditions within much less than 2 mo (Hawkins and Bayne 1992). Therefore, the persistent effect of Origin indicates the importance of endogenous factors, which may include the following influences; (a) genotype; (b) long-lasting environmental effects; (c) parasitism. Although the parentage of mussels used in this work cannot be ascertained, genetic influences seem highly unlikely. The proba- bility that different transplant groups came from different genetic stocks can be assumed to be negligible, because most farmers in the Ri'a de Arousa position their seed ropes on the same area at the rocky shore near the mouth of the Ria. Long-lasting environmental effects mediated through physio- logic changes in early life history have been invoked by Peterson and Beal (1989) to explain origin-dependent variability in the growth rates of Merceiuiria mercenuria . This possibility must not be discarded and is currently being tested in mussels from the Ria de Arousa. Alternatively, parasitization is well known to alter the physio- logical status of bivalves (Newell and Barber 1988). In the Ria de Arousa, variable degrees of infestation by parasites have been reported among cultivated mussels (Figueras et al. 1991). Further. Villalba et al. (1993) examined the degree of infestation of mus- sels by Marleilia refringens. a protistan infecting the stomach and digestive diverticula, during 1988 and 1989 in the raft groupings where our rafts A and C were located. According to their results. there is a gradient, from high prevalences in the inner area (site of our raft C) to lower prevalences in the vicinity of Arousa Island (groupings of rafts A and B). According to Figueras et al. (1991). when invading the stom- ach and digestive diverticula, Marleilia causes tissue disruption in the stomach wall and hemocytic infiltration into the digestive gland. These alterations may, obviously, disturb the normal pro- cess of food adsorption, which could be the reason for lower absorption efficiencies in mussels from origin C compared with A and B. Given that Marteilia does not damage the filtering system, it is possible that lower CR may constitute a response to reduced functional digestive capacity. Other indirect evidence comes from data obtained by Bayne et al. ( 1979), who recorded reduced CR at high temperatures in mussels whose digestive systems were in- fected by the parasite copepod Mytilicola intestinalis. In addition, reduced CR in oysters occur with infection by the endoparasite Haplosporidium nelsoni (Newell 1985). In conclusion, parasitism may help to explain the persistence of observed differences in AE and CR. but this needs future confirmation. The effect of Origin on CI can be attributed to two different sources of variation; First, the condition of mussels before trans- plant took place can account for a certain proportion of observed differences, and second, as a result of the effect of Origin on parameters of energy gain, when exposed to the same environ- mental conditions, mussels from different origins would have ex- perienced different growth rates. We shall deal with this aspect in the next section. Relationship Between CI, Feeding, and Absorption Figure 6 shows CI values for each transplant condition plotted against their corresponding AE. Although these two parameters appear correlated (r" = 0.80; p = 0.001 ) in the pooled set of data, more information can be gained by considering a particular rela- tionship for each different origin. In each of the three cases, CI seems to be an almost linear function of AE, implying that dif- ferences in AE induced by food quality account for a great pro- portion of raft-dependent variability in growth rates. Alternatively, differences in food availability between rafts were of minor rele- vance . Another interesting aspect of the relationships presented in Fig- ure 6 is evident when comparing differences between the slope and the elevation of the fitted lines. The slopes of these lines seem to be related to the mean CR obtained for each Origin; steeper slope appears to be associated with the lowest CR for Origin C. This is as expected, because the degree of dependence of CI on AE must 0.25 0.20 Q Z O 0.15 H 5 z o U 0.10 0.0 0.2 0.4 0.6 0.8 ABSORPTION EFFICIENCY Figure 6. CI plotted as a funtlion of AE (mean values ±95% confi- dence interval). Lines were fitted by eye to points corresponding to the same origin. Circles, Origin A; squares. Origin B; triangles. Origin C. Feeding and Absorption in Transplanted Mussels 679 itself be a function of the rate of feeding, which under the same feeding regimen, can only be a function of the pumping activity. On this basis then, we suggest that the remaining fraction of ori- gin-dependent variability in CI not explained by parameters of energy gain (i.e., differences in elevation of lines) may be attrib- uted to different CI of mussels from each origin before the trans- plantation took place. ACKNOWLEDGMENTS The authors are indebted to the Instituto Espaiiol de Ocean- ografia (La Coruna) for supplying the research vessel "Lura." This work was supported by the Xunta de Galicia (Conselleria de Pesca), under research contract Xunta-EN. LITERATURE CITED Aguirre. M. P. 1979. Biologia del mejillon (Mytilus eiiulis) de cultivo de la Ria de Vigo. Bol. Inst. Esp. Oceanogr. 5:109-159. Bayne, B. L, J. M. Gee, J. T. Davey & C. Scullard. 1979. Physiological responses of Mytilus edulis L. to parasitic infestation by M\tilicola intestinaalis. J. Cons. Int. Explor. Mer .^8:12-17, Bayne, B. L.. A, J. S. Hawkins & E. Navarro. 1987. Feeding and diges- tion by the mussel Mytilus edulis L. (Bivalvia, Mollusca) in mixtures of silt and algal cells at low concentrations. J Exp Mar. Biol. Ecol. 111:1-22. Bayne, B. L., D. W. Klump & K. R. Clarke. 1984. Aspects ot feeding, including estimates of gut residence time, in three mylilid species (Bivalvia, Mollusca) at two contrasting sites in the Cape Peninsula, South Africa. Oecologia {Berl.) 64:26-33. Bayne, B. L. & R. C. Newell. 1983. Physiological energetics of marine molluscs, pp. 407-515. /«; A. S. M. Saleudin & K. M. Wilbur (eds.). The Mollusca. vol 4. Physiology, part 1. Academic Press. New York. Bayne, B. L., P. N. Salkeld & C. M. Worrall. 1983. Reproductive eftort and value in different populations of the marine mussel, Mxlilus edulis L. Ch'cologia (Berl.) 59:18-26. Bncelj, V M. & R. E. Malouf. 1984. Influence of algal and suspended sediment concentrations on the feeding physiology of the hard clam Mercenaria mercenaria. Mar. Biol. 84:155-165. Cabanas, J. M., J. J. Gonzalez, J. Marino, A. Perez & G. Roman. 1979. Estudio del mejillon y de su epifauna en los cultivos Ifotantes de la Ria de Arosa. 111. Observaciones previas sobre la retencion de particulas y la biodeposicion de una halea. Bol. Inst. Esp. Oceanogr. 5:45-50. Conover, R. J. 1966. Assimilation of organic matter by zooplankton. Lim- nol. Oceanogr. 11:338-354. Dickie, L M.. P. R. Boudreau & K. R. Freeman. 1984. Influences of stock and site on growth and mortality in the blue mussel (Mytilus edulis). Can. J. Fish. Aquat. Sci. 41:134-140. Figueras.A. J.,C. F. Jardon&J. R. Caldas. 1991. Diseases and parasites of rafted mussels by (Mytilus galloprovincialis Lmk.): preliminary re- sults. Aquaculture 99:17-33. Hawkins, A. J. S. & B. L. Bayne. 1992. Physiological interrelations, and the regulation of production, pp. 171-222. In: E. Gosling (ed.). The Mussel Mytilus: Ecology, Physiology, Genetics and Culture. Elsevier Science Publishers B. V., Amsterdam. Hawkins, A. J. S.. B. L. Bayne & A. J. Day. 1986. Protein turnover, physiological energetics and heterozygosity in the blue mussel, Mvtilus edulis: the basis of variable age-specific growth. Proc. R. Soc. Lond. B 229:161-176. Hawkins, A. J. S.. P. N. Salkeld, B. L. Bayne, E. Gnaiger & D. M. Lowe. 1985. Feeding and resource allocation in the mus.sel Mytilus edulis: evidence for time-averaged optimization. Mar. Ecol. Prog. Ser. 20:273-287. Hawkins, A. J. S., R. F. M Smith. B. L. Bayne & M. Herat. 1996. Novel observations underlying the fast growth of suspension-feeding shellfish in turbid environments: Mytilus edulis. Mar. Ecol. Prog. Ser. 131:179-190. Hughes, R. N. 1969. A smdy of feeding in Scrohiculariu plana. J Mar. Biol. Assoc. U. K. 49:805-823. Iglesias, J. I. P. & E. Navarro. 1991. Energetics of growth and reproduc- tion in cockles (Cerasloderma edule): seasonal and age-dependent vari- ations. Mar. Biol. 111:359-368. Koehn, R. K. & S. E. Shumway. 1982. A genetic/physiological explana- tion for differential growth rale among individuals of the American oyster, Crassostrea virginica (Gmelin). Mar. Biol. Lett. 3:35^2. Landi'n, A. 1987. Salinidades y temperaturas de la Ria de Arosa (N. W. Espana) a lo largo de un ano. Cuad. Marisq. 9:233-264. Mallet, A. L. & C. E. A. Carver. 1989. Growth, mortality, and secondary production in natural populations of the blue mussel, Mytilus edulis. Can. J. Fish. Aquat. Sci. 46:1 154-1 159. Navarro, E., J. I. P. Iglesias, M. Ortega & X. Larretxea. 1994. The basis for a functional respon.se to variable food quantity and quality in cock- les Cerastoderma edule (L.) (Bivalvia. Cardiidae). Physiol. Zool. bl: 468-496. Navarro, E., J. I. P. Iglesias, A. Perez. U. Labarta & R. Beiras. 1991. The physiological energetics of mussels (Mytilus galloprovincialis Lmk) from different cultivation rafts in the Ria de Arosa (Galicia, N.W, Spain). Aquaculture 94:197-212. Newell, R. I. E. 1985. The influence of environmental factors and the parasite MSX on the physiology of the oyster Crassostrea virginica. J. Shellfish Res. 5:91-95. Newell, R I. E. & J. Barber 1988. A physiological approach to the study of bivalve molluscan diseases. Am. Fish. Soc. Spec. Puhl. 18:269- 280. Okumus, 1. & H. P. Sliding. 1994. Physiological energetics of cultivated mussel (Mytilus edulis) populations in two Scottish west coast sea lochs. Mar. Biol. 119:125-131. Otto. L. 1975. Oceanography of the Ria de Arosa. Meded. Verh. K. Ned. Met. Inst. 96:1-210. Perez, A. & R. Gonzalez 1984. La filtracion del mejillon (Mytilus edulis L. ) en laboralorio. In Actas do Primeiro Seminario de Cencias do Mar, As Ri'as Galegas. Cuademo da Area de Cieneias Mannas, Seminano de Estudos Galegos 1:427^37. Perez, A. & G. Roman 1979. Estudio del mejillon y de su epifauna en los cultivos flotantes de la Ria de Arosa. II. Crecimiento, mortalidad y produccion del mejillon. Bol. Inst. Esp. Oceanogr. 5:22^1. Peterson. C. H. 1983. A concept of quantitative senility: application to Che hard clam. Mercenaria mercenaria (L.)? Oecologia (Berl.) 58:164- 168. Peterson. C. H. 1986. Quantitative allometry of gamete production by Mercenaria mercenaria into old age. Mar. Ecol. Prog. Ser. 29:93-97. Peterson. C. H. & B. F. Beat. 1989. Bivalve growth and higher order interactions: Importance of density, site, and time. Ecology 70:1390- 1404. Smaal. A. C. & M. R. van Stralen. 1990. Average annual growth and condition of mussels as a function of food source. Hydrohiologia 195: 179-188. Theissen. B. F. 1977. Feeding rate of Mytilus edulis L. (Bivalvia) from different parts of Danish waters in water of different turbidity. Ophelia 16:221-232. Thompson, R. J. 1984. Production, reproductive effort, reproductive value and reproductive cost in a population of the blue mussel Mytilus 680 Iglesias et al. adults from a subarctic environment. Mar. Ecol. Prog. Ser. 16:249- 257. Umitia, M. B.. J. I. P. Iglesias, E. Navarro & J. Prou 1996. Feeding and absorption in Cerastoderma edule under environmental conditions in the Bay of Marennes-Oleron (W. France). J. Mar. Biol. Assoc. U. K. 76:431-450. Villalba. A., S. G. Mourelle. M. C. Lopez. M. J. Carballal & C, Aze- vedo. 1993. Marteiliasis affecting cultured mussels Mylilus gallopro- vincialis of Galicia (N. W. Spam). I. Etiology, phases of the infection. and temporal and spatial vanability in prevalence. Dis. Aquat. Org. 16:61-72. Widdows, J.. P. Donkm, P. N. Salkeld. J. J. Cleary. D, M Lowe, S. V. Evans & P. E. Thompson. 1984. Relative importance of environmen- tal factors in determining physiological differences between two pop- ulations of mussels iMytilus edulis). Mar. Ecol. Prog. Ser. 17:33—47. Zar, J. H. 1984. Biostatistical Analysis. 2nd ed. Prentice Hall, Inc. glewood Cliffs, NJ. En- Journal of Shellfish Research. Vol 15. No. 3, 681-684, 1996. FACTORS AFFECTING THE DISTRIBUTION AND ABUNDANCE OF MYTILICOLA ORIENTALIS (COPEPODA) IN THE MUSSEL, MYTILUS TROSSULUS, IN BARKLEY SOUND, B.C. CAMERON P. GOATER' AND AMY E. WEBER Bamfield Marine Skitlon Bamfield. British Columbia VOR IBO. Canada and Department of Biological Sciences University of Alberta. Edmonton Alberta. T6G 2E9, Canada ABSTRACT The copepod. Mxlilicola onemalis. was found in 0-80% of summer-collected samples of mussels (Mxlilus irossulus) from iwo sheltered locations m Barclcy Sound, on the western coast of Vancouver Island. It was absent in mussels from two adjacent, wave-exposed locations. Copepod abundance at infected sites was 2 ± 5. which is comparable to that at other sites along the Pacific Coast where this species has been reported. At the most heavily infected location, copepod abundance was highest in the largest mussels (25-35 mm), particularly those collected near the low-tide mark. Mean abundance increased as average host size increased but was not associated with host density. The restncted regional distribution of this copepod in Barcley Sound (and throughout the Pacific Northwest) may be limited by factors that confine transmission to sheltered, muddy estuaries. Within such sites, copepod abundance is highest in large mussels collected near the low-tide mark. KEY WORDS: copepod. blue mussel, Myiilicola onenlalis. Mytilus Irossulus. marine parasite, population biology INTRODUCTION Mytilicola onentalis (Mori. 1935) is a copepod that infects the intestine of bivalves, especially the mussel, Mytilus trossulus (Gould, 1850), on the Pacific Coast of North America. The spo- radic distribution of this copepod along the coast between southern California and Vancouver Island is thought to reflect the spread of infective stages from imported Japanese oysters, Crassostrea gi- gas (Wilson 1938, Bernard 1969, Bradley and Siebert 1978). Al- though the congener. Mytilicola inlestinalis. has been intensively studied in European mussels (review by Davey 1989), M. orien- latis has received little attention. Presumably, its life cycle is similar to that of M. inlestinalis and involves free-living stages (two naupliar and one infective copepodite) and parasitic adult stages (Gee and Davey 1986). In this study, we document the distribution of M. orientalis in mussels from several localities in Barkley Sound, on the western coast of Vancouver Island, British Columbia. On the most heavily infected mussel bed, we also ex- amine how position on the shore (with respect to the tidal cycle), host size, and host density affect copepod abundance. MATERIALS AND METHODS The taxonomy of smooth-shelled Mytilus is problematic. Pre- viously, most authors considered common mussels in temperate waters to be Mytilus edulis. However. McDonald et al. (1991) used genetic and morphological evidence to show that smooth- shelled Mytilus of the North Pacific are M. trossulus and that M. edulis is primarily restricted to western Europe and eastern North America. Four localities were selected within Barkley Sound and sam- pled between August 4 and 8, 1994. Two localities (Grappler Inlet 'Present address: Depanment of Biological Sciences. University of Leth- bridge. Lethbridge, Alberta TIK 3M4 Canada. and Bamfield Inlet) were sheltered from wave action and had muddy substrata. The other two were at the more wave-exposed mouth of each inlet and had rocky/sandy substrata. Al least 10 size-matched mussels (20-30 mm) were collected haphazardly from each of three sites within each locality. The three sites were separated horizontally along the shoreline by at least 40 m. Mus- sels were returned to the laboratory and were kept alive in flow- through aquaria for a maximum of 24 h before dissection. We sampled intensively at the most heavily infected locality (Grappler Inlet) to examine the effects of tidal emersion, host density, and host size on copepod prevalence and abundance. To examine the effects of tidal emersion, two 60-m transect lines were placed from the high-tide mark to the low-tide mark. The shore was gently sloping, and the tidal height difference between the top and bottom of the transect line was approximately 1 m. Both transect lines were placed perpendicular to the shoreline and were approximately 40 m apart. The first 10 mussels (25-35 mm) en- countered at 0 (high-tide mark). 30. and 60 m (low-tide mark) along the transect line were collected. Mussel density was estimated at Grappler Inlet with 0.25 m^ quadrats. First, two lO-m" sites on the mussel bed were demar- cated with a transect line. The two sites were separated horizon- tally by approximately 5 m. and there were no obvious differences in emersion period between them. Five quadrats were then placed randomly within these two sites following the procedures of Mc- Grorty et al. (1990). All mussels larger than 5 mm within each quadrat were placed in a bag. returned to the laboratory, counted, and measured for maximum linear shell length (in millimeters). Ten mussels (25-35 mm) from each quadrat were dissected for copepod abundance. For each host selected for examination, the maximum shell length was recorded before severing the adductor muscle. The effect of host size on copepod abundance was determined by re- gressing shell length on copepod abundance for all mussels col- lected from the 10 quadrats on Grappler Inlet. A subsample of 10 small hosts (smaller than 25 mm) was haphazardly selected from 681 682 GOATER AND WeBER the total sample (collected from the 10 quadrats) to provide abun- dance data within the size range of mussels on Grappler Inlet. Initially, we examined the entire gastrointestinal tract for im- mature and adult stages of copepods. However, in contrast to Gee and Davey (1986), we did not find immature copepods in the stomach. We therefore limited further examinations to the intes- tine and rectum. These regions of the alimentary tract were com- pressed between two glass Petri dishes and examined for copepods under a dissecting microscope. The effect of tidal exposure on copepod abundance was ana- lyzed by analysis of covariance ( ANCOVA). with size of the host as the covariate. For this analysis, copepod abundance data were log transformed. The effects of host size and host density on copepod abundance were analyzed by correlation. RESULTS Mytilkola orientalis was only present at the two sheltered, muddy locations within Barkley Sound (Table 1). No copepods were found in either of the wave-exposed locations. Prevalence ranged from 0 to 70% for samples of mussels collected from Bamfield Inlet to 10 to 80% for Grappler Inlet. Overall, copepod intensity (not including uninfected hosts) was 2.3 ± 1.2 in the 80 infected mussels collected from these two localities. There was a significant increase in mean copepod abundance with increased distance from the high-tide mark (Table 2; Fig. 1). On average, mussels collected from the low-tide mark had 75% higher mean abundance than those collected from the high-tide mark. In data pooled between the two transects. Scheffe's post- hoc comparison showed that mean abundance differed between samples collected at 0 and 60 m but not between pairs of samples collected at other distances. The density of mussels larger than 5 mm ranged between 16 and 76 mussels/0.25 m" in Grappler Inlet. However, variation in host density was not associated with variation in copepod abun- dance (n = 10, r = 0.288, p > 0.05). but there was a significant positive correlation between copepod abundance and mean mussel size (Fig. 2; n = 10, r = 0.943, p < 0.05). The importance of TABLE 2, Summary statistics of ANOVA for the effect of distance from high-tide line on the abundance of M. orientalis in mussels from Grappler Inlet." Source DF MS F P Covariate 1 0.213 4.678 0.035 Transect 1 0.018 0.397 0.531 Distance 2 0.240 5.295 0.035 Transect x Distance T 0.132 2.916 0.063 Residual 53 0.045 ■■ Main effects were adjusted for the effect of the covariate (log host size). DF: degrees of freedom, MS: mean square. host size is also shown by the moderately significant correlation between copepod abundance and the numbers of mussels within a quadrat that were >30 mm in length (n = 10. r = 0.618. 0.01 < p < 0.05). Mean copepod abundance differed between the two 10-m" areas within Grappler Inlet (Fig. 2: F, t, = 44.9, p < 0.001). This difference in copepod abundance may be due to dif- ferences in position with respect to tidal exposure, to differences in average mussel size (F, ^ = 44.9, p < 0.001), or to a combination of both factors. No copepods were found in mussels smaller than 19 mm (Fig. 3). However, for mussels larger than 19 mm, copepod abundance increased exponentially (log y = -0.83 + 0.71 log x; n = 92, r = 0.41, p < 0.01). Mussels larger than 30 mm made up only 9.5%. of the 358 mussels collected at random from the mudtlat; however, these hosts contained 70.4% of the total number of cope- pods. DISCUSSION The distribution of M. orientalis in mussels from Barcley Sound appears to be restricted to hosts from sheltered, muddy estuaries. Such a restricted regional distribution is similar to re- sults from other reported studies of M. orientalis along the north- em Pacific Coast (Bernard 1969, Bradley and Siebert 1978). The TABLE 1. Prevalence and intensity of A/, orientalis in 25- to 30-mm M. trossulus from four locations in Barkley Sound, Vancouver Island, British Columbia. Locality Bamfield Inlet" Grappler Inlet" Santa Maria^ Dixon Island' Prevalence Mean (±SD) Site" (%) Intensity C 10 10 10 10 10 10 10 10 10 10 K) 27 70 60 20 80 0 1.2 ± 0.4 1.1 ± 0.4 1.5 ± 0.8 3.5 ± 0.7 3.4 ± 2.3 0 0 Range 0-2 0-2 0-3 0-4 0-8 " Letters refer to three sites sampled within each locality. " Sheltered locality. ^ Exposed locality. 0) _ A 0) E " 2.5 2.0 1.5 0 '^■Si m « o is- 1.0 a> c o. ra o 0) u 0.5 0.0 Distance from high-tide mark (m) Figure 1. The relationship between distance along a tidal gradient and mean copepod abundance in mussels collected from Grappler Inlet. M. ORIENTALIS IN MuSSELS 683 0) 0) m 3 E "Si ■o o a. V a o u E 3 C C IS 0) 2.8n 2.6- 2.4- 2.2- 2.0 Area A Area B 16 — r- 18 20 22 24 26 Mean mussel size (mm) within quadrats Figure 2. The relationship between the average size of mussels within 10 0.2?-ni~ quadrats in Grappler Inlet and mean copepod abundance. Five quadrats were placed in one 10-m' area (A) of the mussel bed, and five were placed in an adjacent lO-m" area (B). Symbols indicate means calculated from 10-, 20-, and 30-mm mussels collected from within each quadrat. congener. M. inteslinalis. is also restricted to mussels from shel- tered, muddy estuaries in northern Europe (review by Davey 1989). Several factors may determine this restricted distribution. The colonization potential of infective larvae will play a role, as it does in other parasite/host systems (review by Kennedy 1993). Details of the infection process have been established for M . intestmalis by Gee and Davey (1986). Copepodite larvae can remain infective for up to 35 days, and there is no decline in infectivity for at least the first 6 days after hatching. Gee and Davey (1986) suggest that a photonegative response guides larvae to the substrate, after which the rate of infection is passively determined by a host's field of filtration. Even under expenmental conditions of low temper- ature, relatively low host density, and low copepod density, the transmission of larvae to mussels is remarkably efficient (Gee and Davey 1986). Davey (1989) used such evidence to conclude that the transmission of M. inteslinalis to mussels occurs within local- ized sites on a mussel bed, with few (if any) larvae colonizing from external sources. Thus, for both species, factors associated with the estuarine habitat seem to lead to the restriction of Af\7;7- icola to sheltered, muddy sites. An alternative, but not necessarily independent, explanation for the restricted distribution of M. orienlalis is the sporadic dis- tribution of oyster farms within Barcley Sound. The colonization of copepod larvae from imported Japanese seed oysters to local populations of mussels is frequently cited as the mechanism by which M. ohentalis originated, and dispersed, along the Pacific Coast (Wilson 1938, Bernard 1969, Bradley and Siebert 1978). There is a small commercial oyster farm near the mouth of Grap- pler Inlet that could potentially act as the source of larvae. How- ever, we examined small numbers of oysters and other bivalves at this and other commercial oyster farms in Barcley Sound and none was infected (Weber unpub. data). Our data cannot address the importance of the colonization of mussel beds via imported oysters. However, evidence supporting the traditional view that M. orienlalis originated from Japanese seed oysters is anecdotal (Wilson 1938. Bernard 1969. Bradley and Siebert 1978). It is equally plausible that the transmission biology of M. orienlalis restricts it to those few estuaries in the Pacific Northwest where bivalves (especially M. irossulus) en- counter slow-moving water. In this case, M. irossulus would act as the primary source of larvae, which are then available to infect other bivalves. Ecological studies aimed at investigating transmis- sion between mussels and oysters, together with genetic studies on Japanese and North American stocks of copepods, could address this issue. Within sheltered sites, the abundance of M. orienlalis is influ- enced by local conditions. Host size plays a major role. Similar results have been shown for some field studies of A/, inteslinalis in mussels from Europe, but not all (review by Davey 1989). A positive association between abundance and size is often explained by an accumulation of copepods with the age of the host, an increased tolerance of larger hosts to higher copepod intensities, or to the increased filtration rate of larger hosts. Gee and Davey 's (1986) experimental and field evidence show that the first two possibilities are unlikely for the M. inlestinalis/M. edulis system, but that higher filtration rates were correlated with higher rates of infection. We cannot distinguish among these three possibilities with our data. Whatever the mechanism, it appears that mussels larger than 30 mm, despite their scarcity in Grappler Inlet ( 10% of 358 mussels), are disproportionally important to the overall trans- mission of M. orienlalis within the mussel bed. The significant correlation between the mean size of mussels within 0.25-m" quadrats and mean copepod abundance provides further support for the importance of host size in the local transmission of cope- pods within suitable sites. The difference in mean copepod abun- dance between the two adjacent sites on Grappler Inlet may also be explained by the significant difference in mean host size between the two sites. The location of hosts with respect to the tidal cycle also influ- enced copepod abundance. Similar results have been shown for M. inieslinalis in Europe (Hepper 1955). The most simple explanation for this result is the increased duration that mussels are covered by the tide and thus available to larvae. Alternatively, mussels at lower tidal levels may differ in host quality or there may be fewer 0) (0 (0 3 I o Q. 0> Q. O O O Si E 3 H- y = 0.006x 1-689 6- • 4- • •• • ^^^ 2- • •,.Mli* 0- 5 10 15 20 25 30 35 40 IVIussel size (mm) Figure 3. The relationship between host size and copepod abundance for 92 mussels collected from Grappler Inlet. 684 GOATER AND WeBER large mussels at higher tidal levels to act as significant sources of infection. The infection characteristics of M. oneiUalis in M. trossulus seem similar to those of the more intensively studied system in Europe involving M. intestinalis in M. edulis. In both systems, the regional distribution of copepods in mussels seems to be deter- mined by factors that restrict colonization by free-swimming lar- vae. Such factors are unknown for this system, but wave action, tidal currents, salinity, and/or substratum conditions may each play a role. In both systems, the local distribution of copepods within a mussel bed is influenced by a mussel's position on the shore (m relation to the tidal cycle) and by the size of individual hosts. Because events associated with local transmission between immediate neighbors play an important role in an individual host's risk of infection, larval site selection may be an important feature in determining an individual's ultimate exposure to copepods. Acknowledgments Thanks to T. Goater, J Holmes, and the students and staff of BMS for assistance throughout the project. LITERATURE CITED Bernard. F. R. 1969, The parasitic copepod Mylilicola orienialis in Bntish Columbian bivalves. J. Fish. Res. Bd. Can. 26:190-191 Bradley. W. & A. E. Sieben. 1978. Infection oiOslrea hirida and Mytilus edulis by the parasitic copepod Mylilicolu oneniulis in San Francisco Bay, California. The Veliger 21:131-134 Davey, J. T. 1989. Mylilicola iniesiinalis (Copepoda: Cyclopoida): a ten year survey of infected mussels in a Cornish estuary, 1978-1988. J. Mar. Biol. Assoc. U.K. 69:823-836. Gee, J. M. & J. T. Davey. 1986. Experimental studies on the infestation of Mylilus edulis (L.) by Mylilicola intestinalis Steuer (Copepoda. Cyclopoida). J. Conseil 42:265-271. Hepper, B. T. 1955. Environmental factors governing the infection of mussels Mylilus edulis by Mylilicola iniesiinalis. Fish. Invest. Land. 2:1-21. Kennedy, C. R. 1993. Introductions, spread and colonization of new lo- calities by fish helminth and crustacean parasites in the British Isles: a perspective and appraisal. J. Fish Biol. 43:287-301. McDonald, J, H.. R. Seed, & R. K. Koehn. 1991, Allozymes and mor- phometric characters of three species of Mytilus in the Northern and Southern Hemispheres. Mar. Biol. 111:323-333. McGrorty. S , R. T. Clarke, C. J. Reading, & J. D. Goss-Cuslard. 1990. Population dynamics of the mussel Mylilus edulis: density changes and regulation of the population in the Exe estuary, Devon. Mar. Ecol. Prog. Ser 67:157-169. Wilson. C. B. 1938. A new copepod from Japanese oysters transplanted to the Pacific coast of the United States. J. Wash. Acad. Sci. 28:284- 288. J cmnuil of Shellfish Research. Vol IS. No. 3.685-687. 1996. APPLICATION OF ULTRASOUND TECHNOLOGY TO MOLLUSCAN PHYSIOLOGY: NONINVASIVE MONITORING OF CARDIAC RATE IN THE BLUE MUSSEL, MYTILUS EDULIS LINNAEUS, 1758 PAUL A. HAEFNER, JR.,' BECKY SHEPPARD, JULIE BARTO, ERIN MCNEIL, AND VANESSA CAPPELLINO Rochester Instirute of Technology Department of Biology Rochester. New York 14623 ABSTRACT The noninvasive application of diagnostic medical sonography (ultrasound) to monitor cardiac rate in bivalve molluscs is demonstrated. The cardiac activity of three commercial-size edible blue mussels, Mytilus edulis Linnaeus, was visually monitored within a temperature range from 10 to 20°C in 32%c salinity. Heart rales (17-50 beats/min) were consistent with published studies using various invasive techniques, such as the implantation of electrodes in the pericardial cavity. The acceleration of cardiac rate relative to changes in temperature vaned dunng the course of the observations; acceleration was slower over the temperature range from 10 to 14°C than from 15 to 20°C. KEY WORDS: Cardiovascular physiology, cardiac rate, heart rate, ultrasound, ultrasonography. Mollusca (mollusc, mollusk). Bivalvia (bivalvej, blue mussel, Mytilus edulis INTRODUCTION Galtsoff ( 1964), in his review of the American oyster, referred to the challenges involved with in situ studies of bivalve heart beat. Investigators drilled windows in the valves of oysters and removed the pericardium to make direct visual observations of the heart. This technique was later modified by covering the windows with glass or cellophane. Invasive methods are currently in use, although the current popular technique involves some form of electrode (platinum, silver, stainless steel) inserted through the shell and into the pericardial cavity. The electrodes are connected to an impedance pneumograph with output to a multichannel re- corder (Helm and Trueman 1967, Bayne 1971, Stickle and Sa- bourin 1979, Grace and Gainey 1987. Deaton 1991, Stentin- Dozey and Brown 1994). Depledge and Andersen (1990) developed a noninvasive method for continuously monitoring heart and scaphognathite ac- tivity in rcptant decapod crustaceans. Transducers are cemented to the carapace rather than implanted. Their computer-aided method has also been used successfully with thin-shelled bivalve molluscs and polychaele worms. Instruments used in diagnostic radiology and cardiology have been shown to have potential application to invertebrate organ systems. Gnbble and Reynolds (1993) and Gribble (1994) used angiography to describe the cardiovascular function of a portunid crab, Pornmus pelagicus Linnaeus. Haefner (1996) described the use of diagnostic medical sonography (ultrasound) to monitor and record heart and scaphognathite activity in Portunus gibbesii Stimpson. This article demonstrates the use of noninvasive ultra- sound techniques to monitor heart rate in the edible blue mussel, Mytilus edulis Linnaeus, in response to temperature changes. MATERIALS AND METHODS Edible blue mussels, M. edulis. were obtained from a local vendor (source. Great Eastern Mussel Farms, Inc., Tenants Har- bor, ME). They were maintained in recirculated artificial seawater ' To whom correspondence should be addressed. (32%o, 9°C, 8.1-8.9 ppm dissolved oxygen (DO) in a 50-gallon temperature-controlled system before experimental handling. A depression cast of the ventral half of a mussel was made from Instamold (Activa Products, Inc., Marshall, TX) and secured to a weighted platform, which in turn was placed in a plastic aquarium (26 X 16 X 16 cm) filled with 4.8 L of seawater. The placement of a living mussel in the depression guaranteed that the hinge remained in a position to allow unimpeded ultrasound monitoring of the heart. At least 2 cm of water above the mussel provided a depth range through which the probe could be adjusted to achieve an optimal resolution of ultrasound images. The waterproof transducer probe of the ultrasound unit (Ultra- sonix 750 SDX) contained a 7.5-MHz linear array coupled to a 3-MHz single-crystal Doppler array. Dual video monitors pro- vided a rectilinear image scan of the target organ as well as its frequency of movement. Polaroid prints of the video images were produced on a Sony printer. After the specimen was secured to the platform, the transducer probe was lowered into the water and its surface was cleared of adhering air bubbles. The probe was then positioned parallel to the long axis of the mussel, directly over the dorsal hinge line. This provided the most effective display of the heart, which was de- tected by its known anatomical position (Fig. 1) and by the sono- graphic image of its contractions in the proximity of the posterior adductor and byssus retractor muscles (Fig. 2). The Doppler dis- play of pulse frequency, shown to be effective in monitoring car- diac rate in brachyuran crabs (Haefner 1996), was not consistently reliable in our observations on mussels and was not used. In this study, cardiac rate was determined by counting 10 consecutive contractions, as seen on the video display, and timing their dura- tion with a stopwatch. The rates were converted to beats per minute (bpm). Three mussels (18, 37, and 27 g wet weights, respectively) were monitored at three different temperature regimens in 32%c salinity (Table I). In Trial 1, the water temperature in the exper- imental aquarium was allowed to gradually equilibrate to room temperature. In Trials 2 and 3. the experimental aquarium was placed in a larger basin that served as a water bath. The temper- ature of the water in the experimental aquarium was manipulated by changing the temperature of the water bath with an aquarium 685 686 Haefner et al. CENTIMETERS Figure I. Midsagittal presentation of fresh M. ediilis; left shell, mantle and gills removed. Internal organs identifiable on the ultrasound im- age (see Fig. 2) are posterior adductor muscle (AM), posterior byssus retractor muscle (RM), and pericardial cavity (PC). heater or by adding warm water or ice. The acclimation of the mussels to experimental conditions (15-20 min) was considered complete when the shells gaped, an indication of active ventilation (Bayne 1971) and heart function (Coleman 1974). The extent of gape was visually monitored during the trials, and heart rates were recorded only when the valves were agape. With few exceptions, observations were made at 1-min intervals during the course of the trials. RESULTS AND DISCUSSION We demonstrated that ultrasound technology can be used to monitor cardiac activity in M. edulis. Furthermore, the 17- to 50-bpm range of heart rate for the 10-20°C temperature range (Table I ) is consistent with rates reported by studies that involved invasive techniques (Helm and Trueman 1967. Bayne 1971, Scott and Major 1972, Coleman 1974. Stickle and Sabourin 1979, Grace and Gainey 1987). The metabolic response of M. edulis to increasing temperature, reflected by increased heart rate (Fig. 3), was expected (Coleman 1974). However, there were noticeable differences in the change of heart rate relative to the change in temperature within each trial. Linear regression analyses were performed on subsets of the data (Table I), corresponding to the observed trends shown in Figure 3: Figure 2. Midsagital rectilinear ultra.sound scan of living blue mussel, M. edulis. Compare with Figure 1. Face of transducer probe repre- sented by bright bar across top of photograph. The pericardial space (PC) is located at the confluence of the curved shell (SH), posterior adductor muscle (AM), and the origin of the byssus retractor muscle (RM). The heart, indiscernible in this print, can be detected by its pulsating image in a living specimen. Y = aX + b. where Y is the heart rate (bpm), X is the temperature, a is the slope, and b is the intercept. In Trials I and 2 (Table I ), the increases in heart rate during the first 18- to 20-min period (10-14°C) were relatively gradual, as indicated by a values of 3.35 and 1.42, respectively. Cardiac activity accelerated comparatively faster during periods in which the temperature increased beyond 15°C (Fig. 3, Trials 1 and 2). TABLE 1. Ultrasonographic monitoring of M. edulis in 33%c salinity." Time Temperature Change Cardiac Rate, Linear Regression Range Rate Trial (min) CO (°C/min) Range (bpm) a b r' 1 0-20 13-14 + 0.05 21-26 3.35 -22.38 0.73 21-35 15-18 -t-0.18 27-50 6.80 -72.47 0.89 2 0-18 10-13 -1-0.14 17-21 1.42 2.28 0.91 19-35 14-20 4-0.35 21-35 2.34 -11.96 0.98 36-69 19-12 -0.21 35-20 -2.26 -8.32 0.98 3 0-25 16-18 + 0.07 34-40 1.96 1.44 0.57 26-61 18-10 -0.23 36-24 -2.02 3.76 0.96 " Temperature regimens of trials with corresponding cardiac rates, and linear regression analyses for Y temperature (°C), a — slope of line, b = intercept, and r' = correlation coefficient. aX + b. where Y = heart rate (bpm). X = Ultrasound Detfxtion of Heart Rate in Blue Mussels 687 Temp 00 10 0 20 0 S 60 Jg S *o i ■c n » 20 300 40-0 Temp 0 0 10 0 20 0 30 0 ^011 50 0 60 0 70 0 Time (mm) Figure 3. M. edulis. Heart rates relative to time and temperature change (see Table 1). Trial 1, 18-g specimen exposed to gradual tem- perature increase. Trial 2, 37-g specimen exposed to controlled tem- perature increase and decrease. Trial 3, 27-g specimen exposed to controlled temperature increase and decrease. All three plots are drawn to same scale for comparison. These greater rates of responses are reflected in the higher /> values (6.80 and 2.34, respectively) (Table 1). Lowering temperatures in the latter phases of Trials 2 and 3 resulted in a decrease in cardiac rate. These decelerations were similar in magnitude of slope {b = -2.26 and -2.02) to the accelerations in heart rate (b = 2.34 and 1.96) in response to the increasing temperatures immediately preceding the decrease. During the course of the trials, it was noticed that the width of the gape of the valves increased slightly as water temperature increased. This could be related to an increase in ventilation rate, perhaps in response to decreases in oxygen tension as well as an increase in water temperature. Although ventilation rate and dis- solved oxygen concentrations were not monitored during the course of this study, it is known that changes in oxygen consump- tion are correlated with changes in heart rate (Baynes 1971). As Coleman ( 1974) stated, "Laboratory measurements of heart rate, valve activity, filtration and respiratory rates show that under normal conditions all arc closely related and all would seem to be equally valuable as measurements of activity." The noninvasive ultrasound scanning of bivalves provides conditions that are closer to what Coleman ( 1974) referred to as " normal" than specimens exposed to the implantation of electrodes, thermistors, and trans- ducers in or near the heart. In this article, we demonstrated that the noninvasive monitoring of cardiac activity in M. edulis is possible through the use of ultrasonographic technology. We were also able to relate changes in cardiac activity to observed changes in the behavior of the specimen and/or to changes in the environmental conditions within the test chamber. Currently, there are limitations to the extended application of medical ultrasound technology to invertebrate organisms. Medical ultrasound instruments are designed to distinguish between organs of differing tissue densities and to monitor stroke volume and blood flow in major vessels. In crabs (Haefner 1996) and in the mussel, it was possible to detect and monitor the movement of a soft-tissue organ (heart) lying beneath a relatively dense shell. However, the resolution of the image generated by the instrument in use was insufficient to make any quantitative assessments of stroke volume. The achievement of an optimal transducer fre- quency that will provide enhanced resolution of the heart might soon be possible as the technology of ultrasound imaging systems advances. Unquestionably, limited availability and cost of currently avail- able ultrasound instruments preclude their use in the majority of teaching and research laboratories. However, the ability to com- bine displays of heart activity with direct observation of specimen behavior under noninvasive experimental conditions warrants fur- ther explorations of such application of this technology. ACKNOWLEDGMENTS I am grateful to Michael Foss (Health and Human Services, Springfield Technical Community College, Massachusetts), who encouraged the attempt to apply ultrasonography to invertebrate animals, and to Hamad Ghazle (Director of Medical Sonography. RIT) for present consultation. Bayne. B. L. 1971. Ventilation, the heart beat and oxygen uptake by Mylilus edulis L. in declining oxygen tension. Cnmp. Biochem. Phys- iol. 40A: 1065- 1085. Coleman. N, 1974. The heart rate and activity of bivalve molluscs in their natural habitats. Oceanogr. Mar. Biol. Ann. Hev. I2;30I--^1.V Deaton, L. E. 1991 . Oxygen uptake and heart rate of the clam Pohmesoda earoliniana Bosc in air and in seawaler. J. Exp. Miir. Biol. Ecol. 147:1-7. Depledge, M. H. & B. B. Andersen. 1990. A computer-aided system for continuous, long-term recording of cardiac activity in selected inver- tebrates. Coinp. Biochem. Physiol. 96A;473-477. Galtsoff, P. S. 1964. The American oyster Crassostrea virginica Gmclin Fish. Bull. 64:1-480. Grace, A. L. & L. F. Gainey, Jr. 1987. The effects of copper on the heart rate and filtration rate of Mytilus edulis. Mar. Pollutinn Bull. 18:87- 91. Gribble, N. A. 1994. Static and functional anatomy of the cardiovascular system of the portunid crab Portunus pelagicus Linnaeus. Part A, Static anatomy. J. Crust. Biol. 14:627-640. Gribble, N. A. & K. Reynolds. 1993. Use of angiography to outhne the LITERATURE CITED cardiovascular anatomy of the sand crab Poriunus pelagicus Linnaeus. J. Crust. Biol. 13:627-637. Haefner, P. A., Jr. 1996. Application of ultrasound technology to crusta- cean physiology: monitoring cardiac and scaphognathite rates in Brachyura. Crustaceana 69:788-794. Helm, M. M. & E. R. Trueman. 1967. The effect of exposure on the heart rate of the mussel, Mv/i/Mi edulis L. Comp. Biochem. Physiol. 21:171- 177 Scott. D. M. & C. W Major. 1972. The effect of copper (II) on survival, respiration, and heart rate in the common blue mussel. Mvtilus edulis. Biol. Bull. 143:679-688. Stentin-Dozey, J. M. E. & A. C Brown. 1994. Exposure of the sandy- beach bivalve Donax serra Roding to a heated and chlorinated effluent III. Effects of temperature and chlonne on heart rate. J. Shellfish Res. 13:455-459. Stickle. W. B. & T. D. Sabourin. 1979. Effects of salinity on the respi- ration and heart rate of the common mussel. Mytilus edulis L. , and the black chiton. Kaihcrina lunuata (Wood). J. E.xp. Mar. Biol. Ecol. 41:257-268. Journal of Shellfish Research. Vol. 15, No 3. 689-694. 1996 COMPARISON OF GROWTH, SURVIVAL, AND REPRODUCTIVE SUCCESS OF DIPLOID AND TRIPLOID MERCENARIA MERC EN ARIA* ARNOLD G. EVERSOLE,' CHRISTOPHER J. KEMPTON,' NANCY H. HADLEY,- AND WILLIAM R. BUZZI^ ^ Departmem oj Aquacidlure . Fisheries and Wildlife Clemson University Clemson. South Carolina 'Department of Natural Resources Marine Resources Research Institute Charleston. South Carolina Department of Biology College of Charleston Charleston. South Carolina ABSTRACT Diploid and Iriploid northern quahogs, Mercenaria mercenaria. were cultured intertidally in a South Carolina estuary tor about 4 y. No difference in shell length was detected between quahogs identified as triploids and diploids at 6 and 27 mo of age; however, at 47 mo of age. triploid quahogs were significantly larger than diploids. The survival rates of diploids and triploids were similar over the field growout period. None of the triploids thermally induced to spawn released gametes, whereas 82% of the diploid quahogs released viable gametes. Gonads of diploid individuals were npe, with numerous mature oocytes and radiating bands of spermatozoa in the lumens. Gametogenesis in triploids, however, was severely retarded and abnormal. Although the lumen areas in triploid and diploid quahogs were similar, the areas occupied by sex cells were significantly larger in the diploid quahogs. KEY WORDS: Mercenaria. quahogs. ploidy. growth, survival, reproductive success. INTRODUCTION Since the early 1980s, tnploidy has been induced in a number of bivalves including the eastern oyster, Crassostrea virginica: the Pacific oyster. C. gigcis: the pearl oyster. Pinctada fucata mar- tensii: the bay .scallop. Argopecten irradians: the noble scallop. Chtamys nohilis: the great scallop. Pecten maximiis: the blue mus- sel. Mxtilus edulis: the soft-shelled clam. Mya arenarut: the north- em quahog, Mercenaria mercenaria: and the dwarf surfclam. Mulina lateralis, (Allen et al. 1982, Stanley et al. 1984, Tabarini 1984. Allen and Downing 1986. Beaumont 1986. Buzzi and Manzi 1988. Beaumont and Kelly 1989. Komaru and Wada 1989. Wada et al. 1989. Guo and Allen 1994a). The methodologies and consequences of triploidy induction in molluscan shellfish are re- viewed by Beaumont and Fairbrother ( 1991 ). Although enhanced growth of triploids is a well-established characteristic in many of these species, it has not been documented in the soft-shelled clam (Mason et al. 1988) and the northern quahog (Hidu et al. 1988). In fact, triploid quahogs were significantly smaller than diploid con- trols in Maine waters after three growing seasons (Hidu et al. 1988). Buzzi and Manzi ( 1988) produced triploid quahogs as part of a genetics program in South Carolina. Analysis at 6 mo of age revealed no statistical difference in shell length (SL) between ju- venile quahogs confirmed as diploid and triploid individuals. Qua- hogs reach sexual maturity at 35- to 40-mm SL at about 1-1 .5 y of age in South Carolina (Eversole et al. 1980). The reallocation of metabolic energy from gametogenesis to growth has been pro- posed to explain the increased size in triploids (Allen and Downing 1986). This study was designed to evalute the performance of *Technical Constribution No. 4248 of the South Carolina Agricultural Experiment Station. confirmed diploid and triploid quahogs after several breeding sea- sons. We also histologically examined the gonads and evaluated reproductive potential. The aim of this study was to test the hy- pothesis that triploidy results in increased size (SL), altered ga- metogenesis. and sterility in reproductively mature northern qua- hogs. MATERIALS AND METHODS Production and Growout The quahogs used in this study were produced as a part of a quahog genetics program in South Carolina. In December 1987, Buzzi (1990) used the gametes from three males and one female each to create six unrelated families. Aliquots of eggs from each family were treated with 1 mg/L cytochalasin B in 0.1% dimeth- ylsulfoxide (DMSO) 5 and 10 min after fertilization to disrupt meiosis 1 (M I) and meiosis II (M II). respectively (Buzzi and Manzi 1988. Buzzi 1990). The two experimental groups (MI and M II) and control groups (a DMSO and an unexposed control) were treated for 20 min. After washing, the experimental and control groups were cultured separately in static conditions for 2 1 days before being transferred to recirculating downwelling culture systems for an additional 156 days. In May. at 125 days of age. quahogs were measured and the ploidy was determined by flow cytometry (Allen 1983. Buzzi 1990). The family of quahogs with the highest percentage of triploids (M 1 and M 11) was planted on a private aquaculture lease in Folly River in October 1988. Quahogs from the M 1 and M II treatments were planted in separate trays at about 900/m". In March 1989, quahogs from the M 1 and M II treatments were combined and planted in one tray because of poor survival. This tray was sub- sequently relocated to Charleston Harbor in September 1989 be- cause the aquaculture operation failed. Quahogs were replanted on 689 690 EVERSOLE ET AL. a different aquaculture lease in Folly River in July 1990. where they remained until the termination of this phase of the experiment in October 1991. Tissue Sampling and Ploidy Determination In February 1990, quahogs (n = 483) were measured and individually numbered by etching each valve. A subsample of quahogs (n = 25) representative of the size distribution was se- lected for a test run of the ploidy determinations. The anterior end of the valves was notched with a grinding tool, and a small portion of mantle and muscle was teased free with fine forceps. The tissue was added to a capped vial with 1 mL of Sorensens"s buffer (pH 7.8) until a small pellet was visible on the bottom; 2.3 mL of ethyl alcohol was then added and stirred with a vortex mixer. The sam- ples were held on ice until frozen the following day. Results from a test run of these procedures in February 1990 indicated that ploidy analysis was possible and that the short-term survival of the sampled quahogs in a raceway was good. Tissue samples for ploidy determinations were collected on two other occasions. June and December 1990. A total of 150 ploidy determinations were attempted, and the ploidy was confirmed for all of the quahogs used in histologic examinations, spawing trials, and growth assessments, except in October 1988 and March 1989. Experimental quahogs lacking ploidy confirmation are referred to as either cytochalasin B-treated or control individuals in this pub- lication. The methods for the preparation of tissues and ploidy determi- nation were adapted from Allen (1983) and Buzzi ( 1990). Thawed tissue samples were digested for 30 min at room temperature in 0.5 mL of 0.05% collagenase in 1.5-mL polypropylene centrifuge tubes. Digestion was stopped with the addition of 0.5 mL of phos- phate buffer. The tissues then were aspirated into a 3-ml syringe and repeatedly forced through an 18-gaugc needle to dislodge cells. Cell suspensions were centrifuged for 3 min. and the super- natant was decanted. Cell pellets were digested further in 0.6 mL of a 1;1 solution of 0.003% Nonidet P-40 and 60 (xg/niL ribonu- clease A. After digestion for 30 min at room temperature and stirring with a vortex mixer, the cell suspensions were transferred to a 1-mL tuberculin syringe and forced through a 48-(xm-pore- size Nytex screen into a clean 1.5-mL centrifuge tube. Samples were again centrifuged, decanted, and washed with 0.5 mL of phosphate buffer. One-half hour before flow cytometry, samples were centri- fuged and decanted. 0.6 mL of 50 jjig/mL propidium iodide was added, and the samples were stirred with a vortex mi.xer. Cell preparations were filtered once more immediately before loading in an EPIC 751 flow cytometer equipped with an argon laser. Ploidy determination was accomplished by a comparison of the relative fluorescence of unknown samples with those from known diploids. Growth The SL (anterior-posterior axis) of those quahogs identified as diploids and triploids was measured to the nearest 0. 1 mm in February 1990 and October 1991. These diploids and triploids were treated as a population, and the mean SL of quahogs were compared by the use of paired t-tests with the appropriate statistic for unequal and equal variances. The numbers of identified diploid and triploid individuals available for SL comparisons were 84 and 37 in February and 42 and 15 in October, respectively. Survival Survival was estimated by companng the relative proportion of triploids to diploids determined by Buzzi (1990) when the quahogs were 4—5 mm in SL and approximately 6 mo of age in May 1988 to the relative proportion of triploids alive in February 1990 (ap- prox. 27 mo of age) and October 1991 (47 mo of age). The null hypothesis that the relative proportion of tnploid to diploid qua- hogs was the same in February 1990 and in October 1991 as that determined in May 1988 was tested with X". Spawning Trial and Gonad Conditions In April 1991. 22 diploid and 22 triploid quahogs were col- lected from the growout location and conditioned for a month before the spawning trial. Quahogs in individual vessels were induced to spawn by thermal induction and the introduction of pasteurized sperm (Castagna and Kraeuter 1981 ). The number and sex of spawners were recorded. Eggs collected from individual spawns were quantified with a counting cell . The number of sperm in a spawn was estimated with a spectrophometer and the linear relationship (sperm/mL = [45.4284 x lO*^] absorbance at 610 nm) developed by Bricelj (1979). Subsamples of gametes were mixed to determine viability and competence through 48 h. The diameters of 100 formalin-fixed eggs were estimated with an AlC-2 image analysis system. After the spawning inductions, a subsample of 10 triploid and 10 diploid quahogs was sacrificed for histologic examination. Quahogs were shucked and the foot, mantle, and gill tissue around the gonad were cut away before the gonad was placed in David- son's fixative. The entire gonad was embedded in paraplast, and sections were cut at 7 |j.m progressively through half of the gonad. Sections were stained with hematoxylin and eosin Y (Howard and Smith 1983). Eleven evenly spaced sections from each gonad of five triploid and five diploid quahogs (three females and two males) were used to calculate the percentage of lumen area in the microscope field and the percentage of lumen area occupied by sex cells. This calculation involved estimating the area of the gonad lumen and sex cells in the lumen with a computerized scanning image analyzer (Heffeman and Walker 1989). An attempt was made to select microscope fields representative of the entire gonad section. The number of occytes per square millimeter was mea- sured, and the diameters of oocytes and their nuclei were also measured at this time. Only oocytes with a visible nucleolus were measured. Female and male gonad area and the number oocytes per square millimeter were compared separately between triploid and diploid quahogs with f-tests. Because only one triploid had mature oocytes, diameter measurements of oocytes and nuclei were restricted to one individual of each ploidy and statistical comparisons were not attempted. RESULTS Growth At 6 mo of age (May 1988), there were no statistical differ- ences in SL between identified diploid and triploid quahogs re- sulting from the disruption of either meiosis I (M I) or meiosis II (M II) (Buzzi 1990). Mean SL (±SD) for diploids and M I trip- loids were 5.13 ± 0.40 and 4.83 ± 0.80 mm; values were 5. 14 ± 0.44 and 5. 1 1 ± 0.42 mm for diploids and M II triploids, respec- tively (Fig. 1). Shell measurements of the cytochalasin B-treated quahogs revealed that from May to October 1988, the mean SL Diploid and Triploid Quahogs 691 , FtokJ Grow-out n 50 - Folly R Charleston Harbor Folly R. ^ r^* Hatchery Combined I -c ^^l Folly R. r*- E E _L J_ i Y///A ^N 1 1 3N '// cn c O 20 . "S -C 10 • 0- 1 j j 1 i = +■ A A A 1 :g Figure I. Mean SL and !itandard devjatiun of diploid and triploid northern quahogs cultured in Folly River and Charleston Harbor. SC. The growth of quahogs treated to disrupt meiosis I IM I) and meiosis II (M 11) was monitored over a 47-nio period. Shaded and cross- hatched histograms represent SL of ploidv -confirmed triploid and dip- loid quahogs. whereas open histograms represent SL of the cy tocha- lasin B-treated groups. The asterisk indicates a significant difference, and the closed triangles represent major spawning periods (Eversole et al. 1980). See text for further details. increased to 22.44 ± 3.75 mm (n = 25) in the M 1 tray and to 26.40 ± 1.92 mm (n = 25) in the M II tray (N. Hadley, unpub, data). When the M 1- and M ll-treated groups were combined m March 1989, the quahogs averaged 29.38 ± 3.49 mm SL (n = 50). After an additional year in the field (February 1990) and possibly one spawn by the larger quahogs. the mean SL of iden- tified tripioids was larger (41 .41 ± 7.03 mm, n = 37) than that of identified diploids SL (40.48 ±6.17 mm. n = 84). but this difference was not significant (t = 0.7278. p = 0.468). The change in mean SL between the first (May 1988) and second (February 1990) ploidy determination was 35.35 mm SL for the diploids and 36.42 mm SL for the tripioids. This increase in mean SL was not tested for statistical significance because in May 1988. the ploidy determination involved destructive sampling (Buzzi 1990). A significant difference (t = 2.5212, p = 0.008) in mean SL was detected between those identified tripioids (50.13 ± 5.53 mm. n = 15) and diploids (46.21 ± 5.04 mm, n = 42) in October 1991 after 20 more months of growth and at least two major spawning periods (Fig. 1 ). The increase in mean SL of individuals over these 20 mo was 27% (8.71 mm SL) for the tripioids com- pared with only 14% (5.73 mm SL) for the diploids; this growth difference was significant (t = 3.402, p = 0.007). Sunhal The relative proportion of triploid (M I and M II) to diploid juvenile quahogs was 25.8% in May 1988 (Buzzi 1990). In Feb- ruary 1990, 30.6% (37 of 121) of the ploidy identified quahogs were triploid, and in October 1991. 26.3% (15 of 57) of the quahogs were triploid. We failed to reject the null hypotheses (2 x 3 contingency table; x" = 1.234. p = 0.557) that the proportion of tripioids after 21 and 41 mo of growtout was the same as the 25.8% observed by Buzzi (1990). The percent survival of the tripioids (55.6%) and diploids (56.0%) between February 1990 and October 1991 was nearly identical. Spawning Trial One triploid quahog was misidentified; consequently, one less triploid (n = 21) was exposed to spawning stimuli. Mean SL were similar (t = 0.9558. p = 0.345) for the diploids (47.73 ± 4.55 mm) and tripioids (49.41 ± 6.82 mm) in the spawning trial. None of the quahogs identified as tripioids spawned, whereas 82% (6 males and 12 females I of the diploids spawned. On average, males released 4.55 x 10'' sperm and females released 1.17 x 10''eggs. Formalin-fixed eggs averaged 78.0 (im and were relatively uni- form in size (73-89 \s.m in diameter). Samples of pooled sperm successfully fertilized eggs and produced shelled larvae. Gonad Examination The gonads of identified diploid quahogs were ripe; male go- nads contained several rows of spermatogonia, spermatocytes, and spermatids with radiating bands of mature spermatozoa arranged with heads facing the follicle wall and tails toward the lumen center (Fig. 2a). Two of the five tripioids were tentatively identi- fied as males because these quahogs had darkly stained cells in the lumen (Fig. 2b) that appeared to be products of abnormal sper- matogenesis (Loosanoff 1937). These triploid quahogs also lacked definite oogonia and oocytes. The gonads of diploid females contained free mature oocytes (50-80 |xm in diameter) within the lumen, whereas other oocytes were still attached by a thin peduncle to the follicle wall (Fig. 2c). The vitelline coat was fully developed, and a well-defined nucle- olus within the nucleus was prominent in mature oocytes of dip- loids. In contrast, an occasional large oocyte (60-90 |Jim in diam- eter) was observed in the gonad of a triploid quahog, but in many cases, the lumen was empty (Fig. 2d). Small clusters of what appeared to be proliferating cells were also observed along the follicle wall. These cells were not as darkly stained as those cells observed in the triploid males. Although the gonads of female and male tripioids contained some oogenic and spermatogenic stages in every specimen checked, gametogenesis was greatly retarded and obviously aberrant in the triploid quahogs. Data for the percent lumen area per microscope field indicate that female and male lumen areas were similar in triploid and diploid quahogs (Table 1 ). However, significant differences in the lumen area occupied by oogenic and spermatogonic stages were observed between diploids and tripioids; in diploids, sex cells occupied two to four times as much area as in triploid quahogs. The mean number of oocytes per square millimeter in diploid gonads was also significantly larger than that for triploid gonads. The difference in the oocytes per square millimeter between the diploid and triploid quahogs was about two orders of magnitude. The mean diameter ( ±SD) for nucleated oocytes in a triploid was 74.1 ± 12.4 ia.ni (n = 49), compared with 62.6 ± 9.4 |jim (n = 99) in a diploid quahog. The diameter of triploid oocyte nuclei averaged 40.4 ± 9.5 p.m. compared with 30.3 ± 6.0 (xm for diploid quahogs. DISCUSSION The survival of quahogs treated with cytochalasin B through the larval period was significantly lower than the controls (Buzzi 1990). After the larval period, the survival of the cytochalasin B-treated quahogs was comparable to that of the controls from spat to 6 mo of age (Buzzi 1990). Two lines of evidence indicate that survival was also similar in the older triploid and diploid quahogs; the relative proportion of tripioids to diploids did not change from May 1988, and the percent survival of each ploidy was the same from 27 to 47 mo of age. Comparable survival has been observed for 1-y and older triploid and diploid eastern oys- 692 EVERSOLE ET AL. M V .^ Figure 2. Tissue sections of identified diploid and triploid quahog gonads sacrificed in May 1991 from: a) ripe diploid male, 53.6 mm SL; b) triploid male, 54.7 mm SL; c) ripe diploid female, 55.2 mm SL; and d) triploid female, 47.9 mm SL. Bar = 100 p.m. ters. bay scallops, and soft-shelled clams (Stanley et al. 1984, Tabarini 1984, Allen et al. 1986). In contrast to these observa- tions, the survival of triploid Pacific oysters from spat to 1 y was superior to that of diploids (Allen and Downing 1986). Reduced survival in triploid larvae is commonly observed in molluscs, and it is believed to be caused by exposing the fertilized eggs to harsh induction methods during critical developmental stages (Beaumont and Fairbrother 1991). On the other hand, the reasons for the superior survival of post-set triploids are less obvious. Adult trip- loid Pacific oysters produce fewer gametes and, as a consequence, have less energetic demands for reproduction than do diploid oys- ters (Shpigel et al. 1992). These investigators suggested that trip- TABLE 1. Mean (x) ± standard deviation (SDl of the percentages of lumen area per microscope field and sex cell area per lumen and the number of oocytes per square millimeter of gonad." Female Male Diploid Triploid Diploid Triploid Parameter n X ±SD n \ ±SD n X ±SD n X ±SD Lumen per field (%) Sex cells per lumen (%) Oocyte/mm- 3 3 3 78.0 ± 9.6" 34.4 ± i.O" 133.9 ± 28.8' 3 3 3 69.1 ± 6.5 LS.O ± 5.3 1.5 ± 2.6 2 75.0 ±3.8^ 85.2 ± 3.7' 2 2 59.1 ± 14.8 20.0 ± 0.5 " n represents the number of quahogs used to calculate means. Comparisons were made within category and sex. "t = 1.3349, p = 0.2528. •= t = 1.4647, p = 0.2806. "" t = 5.9774. p = 0.0039. ' t = 24.8668, p = 0.0016. 't = 8.2201, p = 0.0012. Diploid and Triploid Quahogs 693 loid oysters gain an energetic advantage and survive better than do diploids under stressful conditions (e.g., elevated leniperaturc). The production of faster growing and larger individuals via trip- loidy has the potential for increased survival through reduced qua- hog predation (Whetstone and Eversole 1978). In contrast to the observations of Hidu et al. ( 1988). triploid quahogs were significantly larger than the identified diploids in this study. The length of the study and the number of spawning periods experienced by the two study populations differed. For example, quahogs were cultured for four growing seasons in South Carolina and only three growing seasons in Maine (Hidu et al. 1988). It is well known that the quahog grows faster and reaches sexual maturity earlier in the more southerly portion of its geo- graphical range (Ansell 1968. Eversole et al. 1980). By the use of Ansell's ( 1968) growth data and a 35- to 40-mm SL minimum for maturity (Eversole et al. 1980). it is unlikely that quahogs grown in Maine were capable of spawning during the three growing sea- sons of the study of Hidu et al. (1988). In contrast, the South Carolina diploid and triploid quahogs reached 35^0 mm SL after two growing seasons and experienced at least two spawning peri- ods before the final diploid-triploid growth comparisons. It is pos- sible that if the Maine study had been extended to five growing seasons, the growth of triploids would have surpassed that of the diploid controls. The induction of tnploidy retarded gametogenesis in the north- em quahog. The sex cells in the lumen occupied less area in both sexes, and fewer oocytes developed in the females identified as triploids (Table 1). Gametogenesis was reported to be severely retarded in triploid soft-shelled clams, bay scallops, and eastern oyster (Tabarini 1984. Allen etal. 1986. Barber and Mann 1991). and to a much lesser, degree in triploid Pacific oysters and dwarf- surf clams (Allen and Downing 1986. Allen and Downing 1990, Guo and Allen 1994a). Triploid Pacific oysters spawn and in some cases produce viable gametes (Guo and Allen 1994b). Unlike the triploid Pacific oyster, triploid quahogs failed to respond to re- peated spawning stimuli, despite the fact that a few large oocytes were in the gonad of one triploid. It appears, on the basis of the overall abnormal nature of the gonads, the scarcity of sex cells, and the failure to respond to spawning stimuli, that triploid north- em quahogs should be considered sterile. Triploid Pacific oysters produced larger eggs and sperm than did their diploid counterparts (Komaru et al. 1994. Guo and Allen 1994b). Because triploid nuclei contain theoretically 1.5 times more DNA than diploid nuclei, the observation that the oocyte and nuclei diameters of triploid quahogs were larger than those in diploid quahogs was not unexpected. In fact. Child and Watkins (1994) used the optically measured diameters of gill tissue and hemolymph cell nuclei to successfully distinguish between known diploid and triploid Manila clams. Tapes philippinantm . Recently. Guo and Allen (1994a) added polyploid gigantism to heterozygosity (Stanley et al. 1984) and energy reallocation (Allen and Downing 1986) as hypotheses to explain increased size in triploid molluscs. Although our study was not designed to test these hypotheses, our data do lend support for the energy reallo- cation hypothesis, which argues that energy normally destined for gametogenesis is allocated to growth because of partial or com- plete sterility. In this study, the evidence is good that identified triploid quahogs were sterile; triploids failed to respond to spawn- ing stimuli and exhibited signs of abnormal and severely retarded gametogenesis. Also, the difference in diploid and triploid SL was not observed until after quahogs reached sexual maturity (i.e.. 35— fO mm SL) and had opportunity to spawn. Ansell and Lander ( 1967) calculated the losses attributed to spawning in a 40-mm-SL quahog to be about 20-25% of the total energy used for growth. This amount of energy, if diverted into growth, may be sufficient to account for the difference in SL between known diploids and triploids; however, it may not be the only factor contributing to the observed SL difference. The relative importance of heterozygos- ity, energy reallocation, and gigantism in increased triploid size may differ among molluscan taxa; understanding these mecha- nisms in different taxa warrants further study (Guo and Allen 1994a). LITERATURE CITED Allen. S. K-, Jr. 1983. Flow cytometry: assaying experimental fish and shellfish. Aquaculture 33:317-328. Allen, S. K., Jr. & S. L. Downing. 1986. Performance of tnploid Pacific oysters, Crassoslrea gigas (Thunberg). I. Survival, growth, glycogen content, and sexual maturation in yearlings, J. Exp. Mar. Biol. Ecol. 102:197-208. Allen. S. K.. Jr. & S. L. Downing. 1990. Performance of triploid Pacific oysters, Crassostrea gigas: gametogenesis. Can. J. Fi.sh. Aqiial. Sci. 47:1213-1222. Allen. S. K. Jr., P. S. Gagnon & H. Hidu. 1982. Induced tnploidy in the soft-shell clam: cytogenetic and allozymic confirmation. J. Hered. 73:421-428. Allen, S. K.. Jr., H. Hidu & J. G, Stanley. 1986. Abnomial gametoge- nesis and sex ratio in triploid soft-shell clams {Mya arenana). Biol. Bull. 170:198-210. Ansell, A D. 1968. The rate of growth of the hard clam Mercenaria mercenaria (L.) throughout the geographical range. J. Cons. Perm. Inl. Explor. Mar. 31:364-409. Ansell, A. D. & K. F. Lander. 1967. Studies on the hard-shell clam, Venus mercenaria. in British waters. 111. Further observations on the seasonal biochemical cycle and on spawning. J Appl. Ecol. 4:425- 435. Barber, B. J. & R. Mann. 1991. Sterile triploid Crassostrea virginica (Gmelin. 1791) grow faster than diploids but are equally susceptible (o Perkinsus marinus. J Shellfish Res. 10:445^50. Beaumont, A. R. 1986. Genetic aspects of hatchery reanng of the scallop, Pecten maximum (L). Aquaculture 57:99-1 10. Beaumont. A. R. & J. E. Fairbrother. 1991. Ploidy manipulation in mol- luscan shellfish: a review. J. Shellfish Res. 10:1-18. Beaumont. A. R. & R. S. Kelly. 1989. Production and growth of triploid Mylilus edulis larvae. J. Exp. Mar. Biol. Ecol. 132:69-84. Bncelj. M. V. 1979. Fecundity and related aspects of hard clam {.Merce- naria mercenaria) reproduction in Great South Bay. New York. Mas- ters Thesis. State University of New York at Stony Brook. 98 pp. Buzzi. W. R. 1990. Effects of induced polyploidy on the growth and survival of juvenile hard clams Mercenaria mercenaria. Masters The- sis. College of Charleston. South Carolina. 58 pp. Buzzi . B . & J . J . Manzi . 1988. Growth and survival of larval and juvenile polyploid clams. Mercenaria mercenaria. J. Shellfish Res. 7:151 (ab- stract). Castagna. M. & J. N. Kraeutcr. 1981 . Manual for growing the hard clam Mercenaria. Spec. Rep. Appl. Mar. Sci. Ocean Engr. No. 249. Gloucester Point, Virginia. 1 10 pp. Child, A. R. & H. P. Watkins. 1994. A simple method to identify triploid molluscan bivalves by the measurement of cell nucleus diameter. Aquaculture 125:199-204. Eversole, A. G., W. K. Michener & P. J. Eldnge. 1980. Reproductive 694 EVERSOLE ET AL. cycle of Menenaria mercenaria in a South Carolina estuary. Proc. Natl. Shellfish Assoc. 70:22-30. Guo, X. & S. K. Allen, Jr. 1994a. Sex determination and polyploid gi- gantism in the dwarf surfclam {Mulinia lateralis Say). Genetics 138: 1199-1206. Guo, X. & S. K. Allen, Jr. 1994b. Reproductive potential and genetics of triploid Pacific oysters, Crassostrea gigas (Thunbergl. Biol. Bull. 187: 309-318. Heffeman. P. B. & R. L. Walker. 1989. Quantitative image analysis methods for use in histological studies of bivalve reproduction. J. Mollusc. Stud. 55:135-137. Hidu, H., K. M. Mason, S. E. Shumway & S. K. Allen. 1988. Induced triploidy in Mercenaria mercenaria L.: effects on performance in the juveniles. J . Shellfish Res. 7:202 (abstract), Howard, D. W. & C. S. Smith. 1983. Histological techniques for marine bivalve mollusks. NOAA Technical Memorandum NMFS-F/NEC-25. Woods Hole, Massachusetts. 97 pp. Komaru, A., K. Koniski & K. T. Wada. 1994. Ultrastructure of sperma- tozoa from induced triploid Pacific oyster, Crassostrea gigas. Aqua- culture 123:217-222. Komaru. A. & K. T. Wada. 1989. Gametogenesis and growth of induced triploid scallops, Chlamys nobllls. Nippon Susisan Gakkaishi 55:447- 452. Loosanoff, V. L. 1937. Spermatogenesis in the hard-shell clam iVenus mercenaria Linnaeus). Yale J. Biol. Med. 9:437—442. Mason, K. M., S. E. Shumway, S. K. Allen. Jr. & H. Hidu. 1988. In- duced triploidy in the soft-shelled clam Mva arenaria: energetic im- plications. Mar. Biol. 98:519-528. Shpigel. M., B. J. Barber & R. Mann. 1992. Effects of elevated temper- ature on growth, gametogenesis, physiology, and biochemical compo- sition in diploid and triploid Pacific oysters, Crassostrea gigas Thun- berg. J. Exp. Mar. Biol. Ecol. 161:15-25. Stanley, J. G., H. Hidu & S. K. Allen, Jr. 1984. Growth of American oyster increased by polyploidy induced by blocking meiosis 1 but not meiosis II. Aquaculture 37:147-155. Tabanni, C. L. 1984. Induced triploidy in the bay scallop, Argopecten Irraillans. and its effects on growth and gametogenesis. Aquaculture 42:151-160. Wada, K. T , A, Komaru & Y. Uchimura. 1989 Triploid production in the Japanese pearl oyster, Plnctada fiicata manensii. Aquaculture 76: 11-19. Whetstone, J. M. & A. G. Eversole. 1978. Predation on hard clams, Mercenaria mercenaria. by mud crabs, Panopeus herbstll. Proc. Nail. Shellfish. Assoc. 68:42^8. Joiinuil ol Shellfish Research. Veil. 15. No. .^ 695-707. 1996, A CARDIAC CELL LINE FROM MY A ARENARIA (LINNAEUS, 1759) STEPHEN J. KLEINSCHUSTER.' JASON PARENT,' CHARLES W. WALKER,^ AND C. AUSTIN FARLEY^ 'Haskin Shellfish Research Lahoralory Port Norris. New Jersey 08349 ^Department of Zoology Rudman Hall University of New Hampshire Durham. New Hampshire 03824 ^US DOC. NOAA National Marine Fisheries Ser\ice Oxford. Maryland 21654-9724 ABSTRACT Using standard tissue culture techniques, we report here the development of a cardiac cell line from Mya arenaria that exhibits anchorage dependency and an observable, high mitotic coefficient. The cultures can be subcultured while retaining reduced mitotic activity KEY WORDS: M. urcnaria, cell line, marine mollusc, tissue culture, proliferation INTRODUCTION Although it has long been possible to establish in vitro primary, anchorage-dependent cultures of various cells of molluscan origin, demonstration of the sustained, observable proliferation and/or passage of these cells, whether from embryonal or adult tissue, has been limited (Perkins and Menzel 1964. Cecil 1969. Hetrick et al. 1981, Li et al. 1966). Generally, marine molluscan cells, cither from dissociated tissue or cells that have migrated from cxplants, will eventually deteriorate (Bayne 1993). This eventuality can be delayed by lowering incubation temperature and/or reducing nu- trients. The present exception to this generalization is that repre- sented by the established freshwater molluscan cell line developed by Hansen (1979) using embryonal tissue. Recently, Walker et al. (1996) successfully cultured and cryo- preserved malignant hemopoietic cells of Mya arenanu using chemically defined medium m spinner culture. By the use of these techniques, cells doubled every 40-50 h for several generations, but eventually deteriorated. Although presenting many advan- tages, this system does not afford the desirable characteristics of an anchorage-dependent system, in which the direct observation of induced or normal intracellular or intercellular activity can be pur- sued. MATERIALS AND METHODS Preparation of Heart Explants Mature specimens of M. arenaria were obtained from Sandy Hook National Park, NJ, and maintained at the Haskin Research Laboratory, Port Norris, NJ. Clam hearts (ventricles) were dis- sected from clams with normal blood cells and with neoplastic blood cells (determined cytologically). With a stereo-dissecting microscope, the hearts were removed so that contamination from adventitious microorganisms was minimized. Hearts (ventricles) so obtained were rinsed several times in sterile seawater and sub- jected to two 20-min rinses in seawater containing antibiotics (pen- icillin. 200 U/mL; streptomycin, 200 (xg/mL). Specimens were placed in a deep Maximov slide, flooded with sterile seawater containing antibiotics, and minced with a scalpel blade into 1-mm' pieces. After mincing, the pieces were rinsed twice with seawater containing antibiotics to remove extraneous tissue and debris, and culture medium was added just before culture. Culture of Heart Explants Primaria r25 tissue culture flasks (Falcon Labs) were used in all experiments. After dissection and preparation, explants were transferred to the culture flasks by pipette and 3.0 mL of culture medium was added. Incubation was at 15°C under ambient air conditions. Initially, culture medium was changed (50%) weekly. After firm attachment by the explant to the substrate and the ini- tiation of cell migration out of the explant, the medium was changed every 4 days (50%). Medium Preparation During the development of this cell line, many media formu- lations were used and evaluated in an attempt to foster cell main- tenance and proliferation. The most successful medium used (hereafter designated as K-P 58) was composed as follows: sterile glass-distilled deionized H^O, 1,000 mL: MEM Eagle (Earle's salts) with L-glutamine and nonessential amino acids. 4.85 g; CaCK • 2H,0. 1,82 g: KCl. 0.68 g: MgCU ■ 6H,0. 4.36 g: NaCl. 24.26 g: MgSO^ • 7H,0, 3. 16 g; HEPES buffer. 5.0 g; and glucose. 0.5 g. To prepare the final medium, the following components were added at the time of use: fetal bovine serum (FBS) and M. are- naria sterile hemolymph. each 10%; insulin/transferrin/sodium selenite supplement, 2% , Antibiotics routinely added to the culture were penicillin, 100 U/mL, and streptomycin, 100 (xg/mL. The pH of the final medium was adjusted to 7.2 with 1 M NaOH; the final osinolarity was 1 100-1 150 mOsm/kg. Sigma Chemical Co. supplied the MEM (M0643) and the insulin/transferrin/sodium sel- 695 696 Kleinschuster et al. Figure 1. Explant of M. arenaria cardiac tissue, 3 wk in culture. Phase contrast microscopy. Scale bar, 0.12 mm. Figure 2. Cardiac cells of Af. arenaria. 4 wk in culture. Phase contrast microscopv. Scale bar, 0.12 mm. Figure 3. Cardiac cells of A/, arenaria, 4 wk in culture. Phase contrast microscopy. Scale bar, 0.06 mm. enite supplement (11884). Sterile Systems (Hyclone) supplied the FBS. Subculturing Six weeks after culture establishment, proliferation and cell numbers were sufficient to permit passage. After a brief rinse with calcium-magnesium free seawater (CMF), the cells were exposed to a sterile trypsin-CMF seawater solution. During this treatment, the cultures were monitored microscopically to determine the cor- rect exposure time. It was found that an exposure time of 1-3 min at room temperature was ideal. When exposure time was deter- mined microscopically to be sufficient, but before sheets of cells were totally detached, the enzyme solution was carefully decanted and medium with 15% FBS was substituted. The flasks were then agitated to completely free the cell sheets from the substrate, and M. Arenaria Cardiac Cell Line 697 Figures 4-7. Cardiac cells of M. arenaria, 4 wk in culture. Notice many mitotic figures. M, metaphase; D, daughter cells. Phase contrast microscopy. Scale bar, 0.025 mm. (Figures are continued on next page.) 698 Kleinschuster et al. M. Arenaria Cardiac Cell Line 699 Figure 8. Cardiac ttlls of .\/. aniiana. 4 «k in culture. MilDsis cit clitTcrinl ctlh are slinHn in sequence in each photomicrograph. R, reference point; P, prophase; M, metaphase; A, anaphase; D, daughter cells. Phase contrast microscopy. Scale bar, 0.025 mm. (Figure is continued on following pages.) the contents were decanted into fresh cuhure flasks to begin the first passage. RESULTS Within 48 h of culture initiation, most cxplants were firmly adhered to the substrate. The migration of cells from the cardiac tissue began 1-3 days later and continued until the explants were in "monolayer form" (Figs. 1-3). Proliferation and visible mitotic activity became evident within 4 wk (Figs. 4-7). Mitotic activity at 15°C was at a lower level than that observed when the culture medium was changed and the cultures were exposed to room tem- perature for 3^ h. The mitotic coefficient at room temperature was determined to be 0.75%. Figures 8 and 9 document much of this activity with various sequential stages in the mitotic cycle, as well as chromosome movement, shown m cultures that were 4 wk old. It Is important to note that after cell division, the daughter cells immediately reattached to the substrate and remained anchor- age dependent. Mitotic activity was present throughout all areas of the monolayered plaques, with the older parts of the plaque (first migrators at the edge of the plaque) showing as much activity as the younger parts of the plaque (later migrators at the center of the plaque). Very little senility was evident in any of the cultures throughout these experiments, and primary cultures continued to proliferate until passage was indicated. After the passage of primary cardiac cultures, sheets of cells were allowed to remain undisturbed for 3 days, after which fresh medium with 107c FBS was substituted. Shortly thereafter, sheets of cells reattached to the substrate and cellular migration, anchor- age dependency, and proliferation became evident in 1 wk (Fig. 10). Current cultures were in primary culture for 6 wk and have been In first passage for 4 wk as of this writing. The cytological evaluation of cells from both normal hearts and hearts from clams with neoplastic disease revealed no detectable differences with regard to division rates, migration rates, anchor- age dependency, contact inhibition, or response to subculturing. The nuclear/cytoplasmic ratios, chromosome structure, and nu- clear profiles appear to be the same in cultured cardiac cells from both types of clams. DISCUSSION Long-term in vitro cultures of prolific marine molluscan cells in anchorage-dependent form afford many opportunities for inves- tigative exploitation by shellfish researchers. Although, by defi- nition (Merchant et al. 1967), successful subculturing of a primary culture is termed a cell line, longevity of that line is indeterminate. Only when successfully passed many times, with a finite and terminal number of generations, or infinitely and immortally, may a cell line be termed an established cell line (Merchant et al. 1967). Generally, however, only cell lines with appropriate ge- netic aberrations or retrograde embryonic genetic expression ex- hibit immortality. It is yet to be determined how many successful passages the 700 Kleinschuster et al. v^-* M. Arenaria Cardiac Cell Line 701 « •' 702 Kleinschuster et al. ' .«W«Siii..f*iT.!»^r.iiSBti «r ' 4 ^■- ^■• A y';^ -JiT* .•*. *-■ INJ^-V -^i -, *r ■r. .^ M. Arhnakia Cardiac Cell Line 703 .Jf 704 Kleinschuster et al. ..^A*- M. Aren:\ria Cardiac Cell Line 705 Figure 9. Cardiac cells of A/, arenaria, 4 weeks in culture. Mitosis of a single cell is shown in sequence. M, metaphusc; A, anaphase; D, daughter cells. Phase contrast microscopy. Scale bar, 0.025 mm. 706 Kleinschuster et al. ■ r ■ ^ ^H ■ ■ ^^^^^^B^^l ^^^^^H V ^^^ • ■■ •- ^^^ ^^^1 m ^ j' ."^ '} ^m ^^ S- . ■ ■ t I i 12J f^ /■" ■■' tvjj H pr" :^- V 1 FT? * > M ■ Ik id 1^ SHK^^Ci^ ^Qj^ HfH Figure 10. Cardiac cells of M. arenaria, 7 wk in culture, alter the flrst passage of primary culture. Mitosis of a single cell is sliown in sequence. M, metaphase; A, anaphase; D, daughter cells. Phase contrast microscopy. Scale bar, 0.025 mm. cultures described herein will undergo and if they can eventually qualify to be termed an established cell line. On the basis of our observation of reduced mitotic activity after the first passage of the cultures, we expect that the generation number of the cardiac cells therein will be infinite, as are all nomial cells. However, the demonstration of a long-term, anchorage-dependent cell line with an observable high mitotic index fulfills many desirable criteria for exciting studies that heretofore could not be performed. ACKNOWLEDGMENTS This work is identified as Hatch Project No. 32100 and paper No. 0-32100-4-96 by the New Jersey Agricultural Experiment Station; No. 96-22 by the Institute of Marine and Coastal Sci- ences. Rutgers University: and Hatch No. 353 and NIH CAT 1008- 01 by the University of New Hampshire. I LITERATURE CITED Bayne, C. J. 1993. Breaking the barriers to continuous propagation of mol- luscan cells: a summary, pp. 21 . In: Rosenfield (ed. ). Marine Invertebrate Cell Culture: Breaking the Barriers. NOAA Technical Memorandum NMFS-f/NEC-98. Northeast Fishenes Science Center. Woods Hole. MA. Cecil. J. T. 1969. Mitosis in cell culture of the surf clam Spisula solidis- sima. J. Imerlebr. Pathol. 14:407-410. Hansen. E. L. 1979. Initiating a cell line from embryos of the snail Bi- ompluilariii f^hihrala. Tissue Cult. Assoc. Man. 5:1009-1014. M. Arenaria Cardiac Cell Line 707 Hetrick, F. M., E. Stephens. N. Loxmax & K. Lutrell. 1981. Attempls to develop a marine molluscan cell line. Technical Report No. UM-SG- TS-81-06 Maryland Sea Grant Program. University of Maryland, Col- lege Park. MD. Li. M. F..J. E. Stewart & R. E. Drinnan. 1966. /n nrro cultivation ot the cells of the oyster, Crassoslrea virf;inica. J. Fish. Res. Bd. Can. 2.^:.'i95-599. Merchant, D. J., R. H. Kahn & W. H, Murphy. 1967, Handbook of Cell and Organ Culture 2nd ed. Burgess Publishmg Co.. Minneapolis, MN. Perkins, F. O. & R. W. Menzel. 1964 Mamtenance of oyster cells in viiro. Nature 204:1 106-1 107. Walker. C. W.. S. A. Key. J. E. Mulkera. S. Veniia & J. A. Jacobs. 1996. Expression of the tumor suppressor gene p53 in normal and leukemic clam blood cells in vivti and in vitro. J. Shellfish Res. 15:520. Jounuil of Shellfisl, Rcseanh. Vol 15. No. 3. 709-71.3. 1996. GROWTH AND MORTALITY OF TRANSPLANTED JUVENILE HARD CLAMS, MERCENARIA MERCENARIA, IN THE NORTHERN INDIAN RIVER LAGOON, FLORIDA DAN C. MARELLI AND WILLIAM S. ARNOLD Florida Department of Eiivironmeiilal Protection Florida Marine Research Institute 100 8th Aveiute SE St. Petersburg. Florida 33701-5095 ABSTRACT Growth and mortality were examined in hatchery-produced, early-juvenile Mercenaria mercenanu transplanted to protected and unprotected plots at a site in the northern Indian River lagoon. FL. Clam density and size were examined in both treatments five times in the year after transplantation. The growth of clams in both treatments was rapid and comparable to that of clams from other areas within the lagoon. Growth m the protected treatment was initially depressed, but after .363 days, clams from both treatments did not differ significantly in shell height (SH). The mortality of clams in both treatments was high, although significantly greater in the open treatment. Clams in the protected treatment died at a high rate until 80 days into the experiment (SH about 8 mm), beyond which no significant mortality occurred. This experiment suggests that (1) growth rales in the northern Indian River lagoon may favor future aquaculmre ventures; (2) clams can be grown out in the lagoon (if protected from epibenthic predators) when they are 8 mm SH, much smaller than current aquaculture practice suggests; and (3) placing unprotected juvenile clams in situ at high densities is not an efficient stock-enhancement technique KEY WORDS: Growth, mortality. Mercenuria. hard clam, aquaculture. Indian River lagoon INTRODUCTION Growth and mortality in early life stages are important aspects of bivalve population dynamics. The abundance of early-juvenile bivalves is a critical determinant of the abundance of the adult bivalve population (Muus 1973, Marelli 1990) and is negatively affected by mortality. There is also generally an inverse relation- ship between size and mortality that is expressed as a prey size refuge (Carriker 1959. Menzei and Sims 1962. MacKenzie 1977. Whetstone and Eversole 1977, Kraeuter and Castagna 1980, Ar- nold 1984, Peterson et al. 1995). Finally, the livelihood of har- vesters and culturists of commercially important clams depends on the availability or production of adequate numbers of legal-sized clams. The success of bivalve populations is heavily dependent on the survival of postsettlement juveniles, which are the most vulnerable benthic stage (Carriker 1959. Menzei and Sims 1962. Muus 1973, Eldridge et al. 1976, Kraeuter and Castagna 1985). Predation on juvenile clams is often relieved by growth into sizes that offer refuge from predation or by their occupation of a spatial refuge. Spatial refugia occur where physical or biologic structures or phys- iological regimes interfere with predator efficiency (Gainey and Greenberg 1977, Menge 1978, Pohle et al. 1991, Peterson 1982, Summerson and Peterson 1984. Riese 1985, Bertness 1989). Clam culturists construct artificial refugia with predator-exclusion de- vices (Eldridge et al. 1976, Menzei et al. 1976, Flagg and Malouf 1983, Kraeuter and Castagna 1985. Vaughan 1989). We examined three premises regarding the growth and mortal- ity of transplanted juvenile Mercenaria mercenaria (Linnaeus 1758) in Florida's Indian River lagoon: ( 1 ) Clam growth rate in an area historically depauperate of clams is similar to those in areas that support large clam populations; (2) It is economically feasible for clam culturists to begin the field growout phase of their oper- ation with smaller, less expensive clams than those traditionally used; (3) Broadcasting unprotected juvenile clams (I- to 3-mm shell height 1SH|) as a stock-enhancement technique is not effec- tive. MATERIALS AND METHODS Approximately 56,000 hatchery-spawned and hatchery-reared early juveniles, or "seed," of M. mercenaria were held for 1 wk in a 1,600-L conical tank at the Harbor Branch Oceanographic Institution. The tank contained a 62.5 mg/L solution of tetracy- cline hydrochloride, and clams were fed from cultured algae. Wa- ter was not changed for the first 2 days, and subsequently, the water and food supply were changed daily, but no additional tet- racycline was added. Clams were then concentrated on a 750-|im-pore-size screen, and the entire sample population was measured volumetrically. Thirty-two 6-mL subsamples were removed from the sample pop- ulation, and each was placed dry in a glass jar. Jars were trans- ported to the field site in a chilled cooler. A portion of the sample population (approximately 8 mL) was preserved and used to esti- mate the mean size (maximum SH: the maximum measurement from the umbo to the ventral margin) and density of the juvenile clams. The experimental site was located in Shellfish Harvesting Area B ("body B") in the northern Indian River lagoon, just north of State Road 405 on the east side of the Intracoastal Waterway (Fig. I). This area had a depauperate Mercenuria population during 1986 and 1987 (Arnold and Marelli pers. obs.). Hard clam growth rates in body B have been estimated to equal or exceed growth rates from other areas of the lagoon (Arnold et al. 1991). The study site had a sandy bottom, was approximately 2 m deep, and was vegetated with attached and drift algae (Gracilaria sp.). Wa- ter temperatures reach a maximum near 30°C in midsummer and decline to I0-15°C in early winter (Arnold and Marelli pers. obs.). Mean salinity is stable, high (30-36%r; McCall et al. 1970), and similar to that of body C, an area with a large clam population (Arnold et al. 1996). We defined a 2.5- by 2.5-m area on the bottom by laying down a polyvinyl chloride (PVC) template (3/4" schedule 40 pipe) and marking the comers with stainless steel stakes. The template was subdivided with net-mending twine into 16 equal squares (0.39 m~ 709 710 Marelli and Arnold Wabasso i Figure 1. Indian River lagoon, FL, indicating shellflsh-harvesting bodies and approximate position of experimental site (9). each) and was anchored to the substrate during transplanting by four steel rebar pins (9.5-mm |3/8"]). On September 14, 1989, a diver haphazardly poured the clams from one 120-mL jar onto the surface of each of the 16 subplots. The template was then re- moved. A second plot was prepared with a second PVC template and covered on the lower side with polypropylene mesh (open areas. 10 by 10 mm), approximately 4 m away from the first plot. The cage template was also subdivided into 16 0.39-m" squares and was anchored to the substrate by eight rebar pins. The diver planted the clams by pouring the contents of one jar per subplot directly through the mesh onto the substrate. On both treatments, the diver observed that juvenile clams rapidly burrowed into the substrate. Fifteen days after transplantation, three of the subplots from each treatment were sampled. Samplings were also conducted 80. 183, 273. and 363 days after transplanting. The selection of sam- pled subplots was random, but no subplot was sampled more than once during the experiment. Subplots were located by laying a subdivided template over the plot. Three cores with surface areas of 0.032 m" and depths of 5 cm were removed with a suction dredge from each randomly selected subplot. Material removed was collected in a 303-(jLm-pore-size mesh bag and preserved in buffered 10% seawater formalin. After the 80-day sampling, six cores were taken from each subplot because declining densities in the open plots might make statistical analysis difficult. All live M. menenarici that displayed a tetracycline band under ultraviolet illumination were counted, and the SH of each was measured. Most authors report clam size as shell length (SL). The relation- ship between SL and SH was calculated from clams recovered during the early postplanting stages. During each sampling of the caged subplot, the mesh template was cleared of all fouling growth, which was always minimal. Before transplantation and also after the 363-day sampling, three cylindrical cores (37 mm in diameter, 5 cm in length) were taken haphazardly from within both the open and the caged plots. These were analyzed separately for major sedimentary characteristics (% gravel, % sand, % silt-clay, and % organic matter by ignition [Folk 1974]) as a measure of the influence of the treatments on the sediment profile. We analyzed survivorship using a two-way analysis of variance (ANOVA) with days after transplantation and plot condition (open or caged) as factors. Lack of treatment replication may make in- terpreting the meaning of between-treatment effects difficult, but highly significant differences would suggest real main effects. Because the experimental design was unbalanced, data were ana- lyzed with the SAS GLM procedure (SAS Version 5; SAS Insti- tute, Inc., Cary, NC). Where F ratios were significant (p < 0.05), Hochberg's GT2 method for comparing means was applied (Hoch- berg 1974) because it is useful when variances are equal but sam- ple sizes are unequal (Day and Quinn 1989). We used ANOVA to analyze shell data and developed a growth model by fitting (via Table Curve 2D software. Version 3.0; Jandel Scientific. Corte Madera, CA) a nonlinear growth function to the SH data. Separate functions were fit to the data from each treatment because prior ANOVA results demonstrated significant growth differences be- tween treatments. Five functions that have been demonstrated to be useful in modeling bivalve growth (von Bertalanffy, Gompertz, power curve, logistic curve, and exponential curve) (Kennish and Loveland 1980. Kaufmann 1981, Walker and Humphrey 1984, Jones et al. 1990, Arnold et al. 1991. Allison 1994. Lefort 1994) were fit to the SH data. The most appropriate function for each of the data sets was selected on the basis of best fit (highest r value) among the five functions. RESULTS The mean clam density at planting was 5,221.2/m", and the mean SH of these clams was 1.63 ±0.01 mm (range. 0.7-3.7 mm SH). The relationship between SH and SL was determined to be SH = -0.209 + (0.967) (SL) + (-9.127 x 10"^ (SL). Clams in both treatments experienced high mortality almost immediately (Fig. 2): mortality approached 90% within 15 days in the open plot and exceeded 40% in the caged plot. Mortality in both treatments exceeded 95% within 80 days after planting, but clam densities did not decline appreciably for the remainder of the experiment. Be- cause of the drastic early mortality in both treatments, the clam- density data for the 15-day sampling was eliminated from the analysis and the two-way ANOVA was conducted on the remain- ing data. Within-treatment densities on sample dates from 80 to 363 days did not differ significantly from each other; however. clam densities in the caged plot were significantly higher than those in the open plot on all sample dates (p < 0.0001). Sample date had a significant effect on clam height (p = 0), and mean sizes on all dates were significantly different from each other. Treatment also had a significant main effect on clam height (p < 0.0001 ). although there was a significant interaction between date and treatment (p < 0.0001). Clam growth initially appeared to be more rapid in the open plot (Fig. 3). For both treatments, a power curve provided the best fit for clam height over time. The open-treatment data yielded a power curve of the form SH, = 0.021(t + 78.99)' ''°^ where t = days after planting and SH, = Growth and Mortality of Hard Clams 711 UJ cc o o oc LU a. (/) < -I o A 300 200 100 50 10 B E3 E3 E3 E3 015 80 183 273 t (DAYS POST-PLANTING) 363 HI CC O O OC Ol 0. < -I o B 300 200 100 50 H 10 1 - B E3 E3 E3 E3 015 80 183 273 t (DAYS POST-PLANTING) 363 Figure 2. Changes in density (number per 0.032-ni- core) over time of M. mercenaria transplanted to (A) uncaged and (B) caged experimen- tal plots in Indian River shellfish-harvesting body B, 1989-1990. Sym- bols indicate range, mean, and ±1 standard error. Densities v»ere significantly greater in the caged treatment on all dates (p --^ 0.0001). shell height at t (r" = 0.959). The caged treatment shell growth was best expressed by a power curve where SH, = 0.002 (t + 634.24)'^'' (r^ = 0.945). The substrate sedimeni profile was altered by the cage treat- ment. Large increases in both silt-clay (>2227f ) and organic frac- tions (>42%) and a slight ( 10.4%) reduction in the sand fraction were identified in the caged treatment, whereas silt-clay increased only slightly (28.8%) in the open treatment. DISCUSSION Clams transplanted mto Shellfish Harvesting Area B grew at rates consistent with those estimated (from models generated by Arnold et al. 1991 ) for clams in other Indian River areas, rates that could be amenable to economical aquaculturc. Differences in ini- tial clam growth rates observed between caged and uncaged treat- ments were not detectable at the termination of the experiment. Caging artifacts can alter biologic processes, including growth (Vimstein 1977. Dayton and Oliver 1980, Riese 1985). Although we identified an altered sediment profile in the caged treatment, clam growth was ultimately not affected. Mortality rates for unprotected clams were high and consistent with the 96-100% mortality over 3-12 mo reported for juvenile clams transplanted by other researchers (Menzel et al. 1976, Krae- uter and Castagna 1977b. Flagg and Malouf 1983, Kraeuter and Castagna 1985). The rapid decline in clam density followed by a slow but steady reduction in the open treatment is indicative of the density-dependent predation reported in other crab-clam assem- blages (Mansour and Lipcius 1991, Boulding and Hay 1984). Despite the compromise created by a lack of replication, our data suggest that planting unprotected clams of less than 8 mm SH is neither an efficient stock-enhancement technique nor an econom- ical method of aquaculture. In fact, mortality, although reduced in our caged treatment, was unacceptably high (as per Menzel et al. 1976) in either treatment for an aquaculture operation. Several 15 80 183 273 t (DAYS POST-PLANTING) 363 15 80 183 273 t (DAYS POST-PLANTING) 363 Figure 3. Size (SHi over time of M. mercenaria transplanted to (A) uncaged and iBl caged experimental plots in Indian River shellfish- harvesting body B, 1989-1990. Symbols indicate range, mean, and ±1 standard error. Curves represent best fit of models examined and are explained by the equations (a) SH, = 0.021 (t + 78.99)"""' and (b) SH, = 0.002 (t + 634.24)"". Initial growth was higher in open plots (p < 0.0001), but SH did not vary by treatment at t = 363 days. 712 Marelli and Arnold factors may have contributed to high mortality: initial clam sizes were much smaller than those of clams usually transplanted to Indian River field growout facilities (typically, 14—16 mm SL; Barry Moore pers. comm.). and mesh sizes used by aquaculturists in Indian River growout operations are much smaller (commonly 6.35-mm or 1/4" mesh; Barry Moore pers. comm.) than the size we used. Although our cage was not specifically designed to ex- clude crabs tunneling into the treatment, no such behavior was apparent until the end of the experiment, when one stone crab (Menippe mercenaria) had taken up residence under the eastern edge of the mesh. Increasing the initial size of the transplanted clams would re- duce mortality. Caged clams experienced predation below the 10- mm mesh from infaunal or small epibenthic predators. Xanthid crabs are known to enter such cages and prey on juvenile Merce- naria (MacKenzie 1977. Eldridge et al. 1979, Walker 1984, Krae- uter and Castagna 1985. Bisker and Castagna 1989), as are juve- nile blue crabs {CalUnectes sapidus) (Walker 1984, Bisker and Castagna 1989). From an economic perspective, the smallest clams that can be protected and raised should be planted (Kraeuter and Castagna 1977a). Those authors insist that the greater losses of smaller clams can be offset by the lower cost of raising or pur- chasing smaller stock. Survival can be enhanced if protected clams are planted at larger sizes or planted in combmation with predator- exclusion devices and/or predator-removal techniques (Eldridge et al. 1976, Menzel et al. 1976, Whetstone and Eversole 1977, El- dridge et al. 1979, Walker 1984, Kraeuter and Castagna 1985). Some of those authors also reported reductions in predation when clams achieved a SH of 15-20 mm (Menzel et al. 1976. Whetstone and Eversole 1977, Eldridge et al. 1979. Walker 1984). Our pro- tected clams became effectively immune to predation at a SH of 8 mm (SL = 8.55 mml under 10-mm mesh, although similarly sized clams in the open treatment were still vulnerable, suggesting a size refuge from predation by infaunal and small epibenthic predators at about 8 mm SH. Unprotected clams also achieved this refuge, but they continued to be exposed to larger epibenthic pred- ators (brachyuran crabs, busyconid whelks, and fish). Stock-enhancement or aquaculture operations for hard clams must mitigate predatory losses by protecting clams or planting at low densities (see Peterson et al. 1995). The growth and mortality rates we observed may not be directly applicable to clams in other areas, but they do suggest that clams can be economically cultured in Indian River Shellfish Harvesting Area B and, further, that protected clams can be successfully planted at a SH ^8 mm (SL 3=8.55 mm). ACKNOWLEDGMENTS Juvenile clams were provided by Charlie Sembler and Charlie Sembler. Jr., of Sembler and Sembler Seafood, Grant, FL. David Vaughan and Richard Baptiste of Harbor Branch Oceanographic Institution helped label the clams with tetracycline. Clarita Lund and Catherine Bray provided field and laboratory assistance. Barry Moore advised us on Indian River aquaculture practices. LITERATURE CITED Allison, E. H. 1994. Seasonal growth models for great scallops (Pecten nuuimus (L.)) and queen scallops [Aequipeclen opercularis (L.)). J. Shellfish Res. 13:555-564. Arnold, W. S. 1984. The effects of prey size, predator size, and sediment composition on the rate of predation of the blue crab. Callinecles sapidus Rathbun. on the hard clam Mercenaria mercenaria (Linne) 7. Exp. Mar. Biol. Ecol. 80:207-219. Arnold. W. S.. T. M. Bert. D. C. Marelli. H. Cruz-Lopez & P. A. Gill. 1996. Genotype-specific growth of hard clams (genus Mercenaria) in a hybrid zone: variation among habitats. Mar. Biol. 125:129-139. Arnold. W. S.. D. C. Marelli. T. M. Bert. D. S. Jones & I. R. Quitmyer. 1991. Habitat-specific growth of hard clams Mercenaria mercenaria (L.) from the Indian River. Florida. J. Exp. Mar. Biol. Ecol. 147:245- 265. Bertness. M. D. 1989. Intraspecitlc competition and facilitation in a north- em acom barnacle population. Ecology 70:257-268. Bisker. R. & M. Castagna. 1989. Biological control of crab predation on hard clams Mercenaria mercenaria (Linnaeus. 1758) by the toadfish Opsanus tail (Linnaeus) in tray cultures. J. Shellfish Res. 8:33-36. Boulding. E. G. & T. K. Hay. 1984. Crab response to prey density can result in density-dependent mortality of clams. Can. J. Fish. Aqual. Sci. 41:521-525. Cairiker. M. R. 1959. The role of physical and biological factors in the culture of Crassostrea and Mercenaria in a sail- water pond Ecol. Monogr. 29:219-266. Day. R. W. & G. P. Quinn. 1989. Comparisons of treatments after an analysis of variance in ecology. Ecol. Monogr. 59:433—163. Dayton. P. K. & J. S. Oliver. 1980. An evaluation of experimental anal- yses of population and community patterns in benthic marine environ- ments, pp. 93-120. In: K. R. Tenore and B. C. CouU. (eds). Marine Benthic Dynamics. University of South Carolina Press, Columbia. Eldridge, P. J.. A. G. Eversole & J. M. Whetstone. 1979. Comparative survival and growth rates of hard clams Mercenaria mercenaria. planted in trays sublidally and intertidally at varying densities in a South Carolina estuary. Proc. Natl. Shellfish Assoc. 69:30-39. Eldndge, P. J., W. Waltz, R. C. Gracy & H. H. Hunt. 1976. Growth and mortality rates of hatchery seed clams. Mercenaria mercenaria. Proc. Nail. Shellfish Assoc. 66:13-20. Flagg. P. J. & R. E. Malouf. 1983. Experimental plantings of juveniles of the hard clam Mercenaria mercenaria. J Shellfish Res. 3:19-27. Folk. R. L. 1974. Petrology of Sedimentary Rocks. Hemphill Publishing, Austin. TX. 184 pp. Gainey, L. F., Jr. & M. J. Greenberg. 1977. Physiological basis of the species abundance-salinity relationship in molluscs: a speculation. Mar. Biol. 40:41^9. Hochberg. Y. 1974. Some conservative generalizations of the T-method in smiultaneous inference. J Miiltivar. Anal. 4:224-234. Jones, D. S., 1. R Quitmyer, W. S Arnold & D. C. Marelli. 1990. An- nual shell banding, age, and growth rate of hard clams {Mercenaria spp.) from Florida. J. Shellfish Res. 9:215-225. Kaufmann. K. W. 1981. Fitting and using growth curves. Oecologia 49: 293-299. Kennish. M. J. & R. E. Loveland. 1980. Growth models of the northern quahog. Mercenaria mercenaria (Linne), Proc. Natl. Shellfish Assoc. 70:2,^0-239. Kraeuter, J. N. & M. Castagna. 1977a. An analysis of gravel, pens, crab traps and current baffles as protection for juvenile hard clams {Merce- naria mercenaria). Proceedings of the 8th Annual Meeting of the World Maricullure Society. Louisiana State University. Baton Rouge, LA. pp. 581-592. Kraeuter. J. N. & M. Castagna, 1977b. Mercenaria culture using stone aggregate for predator protection. Proc. Natl. Shellfish Assoc. 67:1-6. Kraeuter, J. N. & M. Castagna. 1980. Effects of large predators on the field culture of the hard clam. Mercenaria mercenaria. Fish. Bull. 78:538-541. Kraeuter, J. N. & M. Castagna. 1985. The effects of seed size, shell bags. Growth and Mortality of Hard Clams 713 crab traps, and netting on the survival of the northern hard clam Mer- cenaria mercenana (Linne), J. Shellfish Res 5:69-72. Lefort, Y. 1994. Growth and mortality of the tropical scallops; Aimach- Uimvs ficihellaia (Bemardi). Comptopallnim nuhila (Linne) and Mi- imicliliiniys glonosa (Reeve) in Southwest Lagoon of New Caledonia. J. Shellfish Res. 13:539-546. Linnaeus, C. 1758. Systema naturae per regna tria naturae Kith rev. ed. vol. 1. Regnum animale. Stockholm, Sweden. 824 pp MacKenzie, C. L.. Jr. 1977. Predation on hard clam {Mercemiria merce- nciria) populalicins. Trans. Am. Fish. Soc. 106:530-537. Mansour, R. A. & R. N. Lipcius. 1991. Density-dependent foraging and mutual interference in blue crabs preying upon infaunal clams. Mai-. Ecol. Prog. Ser. 72:239-246. Marelli. D. C. 1990. Recruitment of the estuanne sofl-hottom bivalve Pohmesoda caroliniaita and its influence on the vertical distribution of adults. Veliger 33:222-229. McCall, D., J. G. Cook, J. A. Lasater & T. A. Nevin. 1970. A survey of salinity levels in the Indian River-Banana River complex. Bull. Envi- ron. Conlam. Toxicol. 5:414—421. Menge, B. A. 1978. Predation intensity in a rocky intertldal community: effect of an algal canopy, wave action and desiccation on predator feeding rates. Oecologia 34:17-35. Menzel, R. W., E. W. Cake, M. L. Haines, R. E. Manm & L. A. Olsen. 1976. Clam manculture in northwest Flonda: field study on predation. Proc. Nail. Shellfish Assoc. 65:59-62. Menzel. R. W. & H. W. Sims. 1962. Expenmental farming of hard clams, Mercenaria mercenaria. in Florida. Proc. Nail. Shellfish As- soc. 53:103-109. Muus, K. 1973. Settling, growth and mortality of young bivalves in the 0resund. Ophelia 12:79-116. Peterson, C. H. 1982. Clam predation by whelks (Bimvcoh ^PP): exper- imental tests of the importance of prey size, prey density, and seagrass cover. Mar. Biol. 66:159-170. Peterson, C H., H. C. Summerson & J. Huber 1995. Replenishment of hard clam stocks using hatchery seed: combined importance of bottom type, seed size, planting season, and density. J. Shellfish Res. 14:293- 300. Pohle, D. G., V M Bricelj & Z. Garcia-Esquivel. I99I. The eelgrass canopy: an above-bottom refuge from benthic predators for juvenile bay scallops. Argopeclen irradians. Mar. Ecol. Prog. Ser. 74:47-59. Riese, K. 1985. Predator control in marine tidal sediments, pp. 311-321. In: P. E. Gibbs (ed.). Proceedings of the 19th European Manne Biol- ogy Symposium, Plymouth, 1984. Cambndge University Press, Cam- bndge. Summerson, H. C. & C. H. Peterson. 1984. Role of predation in orga- nizing benthic communities of a temperate-zone seagrass bed. Mar. Ecol. Prog. Ser. 15:63-77, Vaughan, D. E. 1989. Clam culture: state of the art in Flonda. J. Shellfish Res. 7:546. Vimstein, R. W. 1977. The importance of predation by crabs and fishes on benthic infauna in Chesapeake Bay. Ecology 58:1 199-1217. Walker, R L. 1984. Effects of density and sampling time on the growth of the hard clam, Mercenaria mercenaria. planted in predator-free cages in coastal Georgia. Naulilus 98: 1 14— 1 19 Walker. R. L. & C. M. Humphrey, 1984. Growth and survival of the northern hard clam Mercenana mercenaria (Linne) from Georgia, Vir- ginia, and Massachusetts in coastal waters of Georgia. J . Shellfish Res. 4:12-5-129. Whetstone. J. M. & A. G. Eversole. 1977. Predation on hard clams, Mercenaria mercenaria. by mud crabs. Panopeiis herhslii. Proc. Nail. Shellfish Assoc. 68:42^8. Journal of Shellfish Research. Vol. 15, No. 3. 715-718. 1996. THE EFFECTS OF LARVAL STOCKING DENSITY ON GROWTH, SURVIVAL, AND DEVELOPMENT OF LABORATORY-REARED SPISULA SOLIDISSIMA SIMILIS (SAY, 1822) DORSET H. HURLEY AND RANDAL L. WALKER Shellfish Aqiuuulture Laboratory University of Georgia Marine Extension Service 20 Ocean Science Circle Savannah. Georgia 3141 1-101 1 ABSTRACT The optimal larval stocking density was determined for laboratory-reared Spisula solitlissima similis (Say). Stocking density treatments of 10, 20, 30, and 50 larvae/niL were analyzed for effects on survival, growth, and development. Twenty-four- hoiir-old larvae were stocked at the above densities in 500-niL tlasks containing seawater at 25 ppt and 2(I°C, All treatments received a daily food ration of 100,000 cells/mL of Tahitian strain ls 25-cm area and 20-cm depth were dug out and sieved through 1-mm-pore-size mesh to collect animals smaller than 5 mm shell length (SL). Sea surface temperature was re- corded at each sampling. Maximum length on the anterior-postenor axis (SL) was re- corded from all individuals with vernier calipers. Subsamples of 30-40 individuals were taken from the monthly samples, and total wet weight (shell and body wet weight) was recorded. All soft parts were removed and dried at 60°C to constant weight to de- termine shell free dry weights (SFDW). Ash free dry weight (AFDW) was obtained by the ignition of dried soft parts at 550°C for 5 h. Valves were also used to age aniinals from external shell ring readings. Body weight cycles and the production ofthe pop- ulation were calculated for the period April 1991 to March 1992 from pooled monthly length-frequency distributions. Body Weight Cycles as an Indicator of Spawning Cycles The parameters of the relationship between SL and SFDW (Eq. 1 ) were estimated by linear regression analysis on log-transformed data. Annual weight cycles for a standard individual of 50 g total wet weight were calculated as SFDW = a SL'^ (1) 'Contribution no. 1111 of the Alfred-Wcgener-lnstitut fiir Polar und Meeresforschung. where SFDW is in grams and SL is in millimeters. Analysis of variance was used to test for significant differences (p < 0.05) of regression lines of the length-weight relationships between suc- cessive months. Body weight cycles were used to identify the reproductive cy- cle, with a decrease in weights between two successive months indicating a spawning event. Gonad histology was not available to 719 720 Urban 73°57' JS'SB' L. L, + (U_ - L,)(l - e" •") (3) '36°31' I 36°32' Figure I. Lucation of sampling slaiidiis in Iht Ba> ol Dichato, Chile. corroborate the indirect evidence of the spawning cycle. However, Urban and Campos (1994) found that the changes in the body weight cycles for all three of their species studied (G. solida. S. solida. P. thaca). were a good indicator of the spawning cycle on the basis of the histologic information available for G. solida. Growth Rale Estimates Annual shell growth rings on the surface of the valves of 30-40 individuals (of each species) were measured with vernier calipers. Disturbance rings were easily distinguished from annual rings be- cause of the stronger and clearer appearance of the latter. Two data sets were obtained: ( 1 ) Individual growth increment data arranged as tagging-recapture data with constant time intervals (t = 1 y). and (2) mean length for each age was calculated from all individ- uals to obtain age-length data. The data were fit to the von Ber- talanffy growth function, VBGF (von Bertalanffy 1938); L, = U(l -K(l-I(,l (2) where L^ is the asymptotic length (in millimeters), K is the growth constant (per year), t is the age (in years), and to is the age at zero length. Growth parameters were estimated by the use of Fabens" method (Fabens 1965). by fitting a rearranged function of Eq. 2 to size-mcrement data pairs (i.e., tagging-recapture data) by an iter- ative nonlinear least-square method (Simplex algorithm; Press et al. 1986); where L, is the length at the beginning and L, is the length at the end of the time interval t, - t, t(, was estimated by fitting the VBGF (Eq. 2) to age-length data with the Simplex algorithm. In large individuals, it is difficult to separate and count the last few growth rings because they are very close together or overlap- ping because of reduced growth. Therefore, it is not unusual to obtain data that underestimate L, and overestimate K because these two parameters are inversely related (Pauly 1979). These parameters were verified by an alternative method (modified Wetherall method) that estimates L^. from length-frequency data (Wetherall 1986, modified by Pauly 1986), where Beverton and Holt"s (1956) Z-equation based on length data is rearranged to a linear aggression equation; L - L' = a -h b L'; Z/K = -(1 -h b)/b; U = -a/b (4) L is the mean length of individuals of length L' and longer, and L' is the length for which all individuals of that length and longer are under full exploitation. In order to obtain correct growth pa- rameters, the growth routine was rerun with fixed values of L„, calculated as above. Mortality Rate Estimates Total mortality Z was calculated with the single negative ex- ponential model; N, = N^e-^' (5) (where t is the time and N^ is the number of individuals at t = 0) and the length converted catch curve (Pauly 1983). Thereby, with the parameters of the VBGF, lengths of pooled length-frequency samples are converted into ages by the use of Eq. 6; (N,/At,) = N,/ e"^"' (6) where N, is the number of individuals in length class i. At, is the time required to grow through this size class, and t, is the age of the middle-length class i. Mortality Z is calculated by linear re- gression analysis; Ln(N|/At,) = a -(- b t,; Z = -b (7) Natural mortality M was estimated on an empirical basis with the relationship between the P/B ratio and maximum age, A,^^^ (Eq. 9l. Allen ( 1971 ) showed that in a steady-state population, the somatic P/B ratio equals Z, if mortality can be described by the single negative exponential model and growth follows the von Bertalanffy growth model. Thus, M can be estimated from em- pirical relations between the P/B ratio and maximum age, be- cause in unexploited populations, Z equals M. Maximum age, Amax- was calculated with the inverse VBGF (Eq. 8); the VBGF parameters and maximum length. L^,^^, were taken from the growth estimates and pooled length-frequency samples. A„,ax = to - 1/K Ln(l - L^JL^) (8) Natural mortality was estimated with two empirical relationships (Eq. 9) taken from the literature; Hoenig ( 1983) and Etim and Brey (1994). The mean value was taken from these estimates. Log(P/B) = a -(- b Log(A„„) (9) Fishing mortality F (per year) was calculated after Eq. 10, and exploitation rate E (per year) was calculated after Eq. 11; F = Z - M (10) E = F/Z (11) Population Dynamics of Bivalves 721 Production Estimates On the basis of the results of the annual body weight cycle, the gonad production of the population P^,,,, (AFDW, in grams per square meter per year) was calculated from length-weight relations (Eq. I ) of the lowest body weight before and the highest body weight during the spawning season from monthly samples: :i: N, (w Igon du w Igon hfl ' (12) where Wj^^.^ j^^ and Wj^^^^ ^.^f are body weights during and before the spawning season in length class i. The length of first maturity was calculated after Eversole (1989) as 259^^ of the maximum length (L„,,J present in the pooled length-frequency data. The somatic production of the population P (AFDW, in grams per square meter per year) was calculated by the weight-specific growth rate method (Crisp 1984; Eq. 13) from the mean of quan- titative samples, pooled length-frequency data, the VBGF- parameters, and the length-weight relation: 1 N, W, G, (13) where N, is the mean number of individuals (N per square meter). W, is the mean body weight (AFDW, in grams) in length class i, and G, (per year) is the weight-specific growth rate: G, b K HLJL,) I) (14) where b is the exponent of the length- weight relation (Eq. 1 ), L, and K are VBGF parameters, and L, is the mean length in length class i. Individual somatic production P,„j (grams of AFDW per square meter per year) was calculated as follows: 1 W, G, (15) and the mean biomass of the population B (grams of AFDW per square meter per year) was calculated as: B = S N, W, (16) The P/B ratio of the population was calculated from somatic production P and mean biomass ©B. RESULTS Reproduction Figure 2A shows the reproductive cycle of G. solida based on histologic sections from Urban and Campos ( 1994), and Figure 2B and C overlay the SFDW and sea surface temperature cycles for G. solida, S. solida, and P. ihaca. Comparing the reproductive cycle (Fig. 2A) with the body weight cycle (Fig. 2B) of G. solida reveals that the high body weights in summer are principally caused by reproductive activities possibly triggered by low tem- peratures in winter (Urban and Campos 1994). Figure 2D and E show the overlay of SFDW and sea surface temperature cycles for the bivalves V. anliqua. T. dombeii, and E. macha from this study. As observed in Figure 2B and C. the general body weight cycle follows the temperature cycle. Tem- peratures and body weights begin to rise from their lowest values in September/October and reach their highest values in January (the highest body weight of V . anliqua was observed in February; that of £. macha was seen in March). All three species had non- significant body weight changes in winter (May/June and August/ September). It can therefore be assumed that the reproductive periods of V. anliqua. T. domheii, and E. macha follow a pattern similar to that observed by Urban and Campos ( 1994) and that on the basis of the body weight cycles, the gonad production of the Frequency [%] IOOt I I developing I § ripe H spent 1 r^ developing 2 [_] spent 2 £ ■S < ^ 1992 Figure 2. (A) Reproductive cycle, based on histologic sections of G. solida. for the period from June 1991 to May 1992. (B and C) Overlay of SFDW cycle for a 50-g total wet weight standard individual and the sea surface temperature (SST) cycle for the period from April 1991 to March 1992 for three Chilean bivalves, G. solida. S. solida, and P. Ihaca. (Fig. 2A-C taken from Urban and Campos 1994). (D and E) As for Figure 2B and C, but for the Chilean bivalves V. anliqua (D) and T. dombeii and E. macha (E). Solid lines (weight cycles): signiricant differences (p "S 0.05) between successive months; broken lines: non- significant differences. population can be estimated as the difference between the lowest annual body weight in winter and the highest body weight during the summer period by linear regression analysis for each class, accordingly. The linear regression parameters of the length weight relationships for the lowest and highest body weights are given in Table 1. 722 Urban TABLE 1. Parameters of the SL (mml — SFDW (g) relationship of the Chilean bivalves V. antiqua, T. dombeii. and E. macha before and during the spawning season." Species Month a V. antiqua T. dombeii E. macha Oct 1991 Feb 1992 Nov 1991 Feb 1992 Oct 1991 Mar 1992 -4.167 -3.839 -4.688 -5.135 -5.725 -4.893 2.590 2.748 2.621 3.048 2.888 2.793 0.77 0.66 0.97 0.98 0.88 0.68 30 30 36 35 30 29 Log(SFDW) = a -I- b Log(SL). Growth The asymptotic lengths. L^, of T. dombeii and E. macha es- timated with tagging-recapture and age-length data (Table 2) are much lower than the maximum lengths. L^^^^ (Table 3), from the pooled length-frequency distributions, which indicates that L^ is underestimated, and accordingly. K is overestimated. The L.^ val- ues estimated with the Wetherall method (Table 2) are therefore better estimates of the parameter. The VBGF parameters K and t,, shown in Table 2 were all estimated with the fixed L^ of the Wetherall method. The corresponding growth curves based on the mean K and t,, values from tagging-recapture and age-length data are plotted together with age-length data in Figure 3. The curve fitting for V. antiqua and T. dombeii is better than that for E. macha. Mortality Length converted catch curves are shown in Figure 4. Total mortality Z values obtained from these regressions are given in Table 3; values are about Z = 1/y. Maximum age (A,,,^^) and natural mortality values (M) of T. dombeii and E. macha are very similar; A^^,^ ~ 14 y, M = 0.3/y (Table 3). The natural mortality of V. antiqua was not computable because L„,^^ > L^. In this case, maximum age, A^^^. cannot be calculated from L^^^ with the inverse VBGF (Eq. 8). TABLE 3. Total mortality Z, estimated with the catch curve-method and natural mortality M, estimated from two empirical relationships (between maximum age and the P/B ratio), of three Chilean bivalves, V . antiqua. T. dombeii, and E. macha.' Species L, M" M" C>M Venus antiqua 1 .084 74 Not computable L^^^ > L^ Tagelus dombeii 0.839 86 14.7 0.230 0.301 0.266 Ensi.'i macha 1.089 180 13.5 0.254 0.328 0.291 ' VBGF parameters, necessary for the catch curve-method and to calculate maximum age. A^^^. were taken from Figure 3. Maximum length, L^^^ used to calculate A^^,, was taken from the pooled length-frequency dis- tribution, a, estimated with the empincal relationship of Hocnig (1983): Log(P/B) = 0.625 -I- 0.982 Log(A„,^J; b, eshmated with the empincal relationship of Etim and Brey (1994): Log(P/B) = 0.682 -t- 1.130 Log(A^^J. Production The parameters of the linear regression analysis of the length- weight relationship used for the calculation of somatic production are given in Table 4. Individual somatic production curves (Fig. 5) have a similar pattern, in V. antiqua. they reach their highest value with 0.50 g of AFDW per individual per year at 45 mm SL and decrease thereafter, whereas for T. dombeii. it is 0.35 g at 60 mm, and for E. macha. it is 0.90 g at 130 mm. Individual gonad production is reflected by exponential curves beginning at 20 mm SL in V. antiqua. 25 mm in T. dombeii. and 45 mm in E. macha. The length-frequency distributions for the three species show that they are normally distributed (Fig. 6). In V. antiqua. the distribution is unimodal, with a size range from 45 to 80 mm, with no small specimens or recruits present. T. dombeii and E. macha each have a major mode with two smaller modes (7". dombeii. between 15 and 50 mm; E. macha, between 45 and 105 mm). The results for the somatic production of the population and mean biomass estimates are given in Table 5. The highest values were recorded for V. antiqua (22.0 and 122.0 g of AFDW/m" per year), followed by E. macha (9.7 and 43.6 g of AFDW/m' per TABLE 2. VBGF parameters of three Chilean bivalves, V. antiqua. T. dombeii, and E. macha, estimated from two data sets: tagging-recapture data (with Fabens method) and age-length data (with the VBGF)." Species L^ K to r^ L." K" t '■ 'o r^" n Tagging-recapture data V. antiqua 74.6 0.219 — 0.99 73.9 0.223 — 0.99 35 T. dombeii 83.5 0.275 — 0.99 88.5 0.242 — 0.99 44 E. macha 163.8 0.321 — 0.99 189.9 0.222 — 0.97 51 Age-length data V. antiqua 73.6 0.225 0 0.99 73.9 0.212 -0.161 0.99 6 T. dombeii 81.0 0.323 0 0.99 88.5 0.221 -0.658 0.97 10 E. macha 163.7 0.318 0 0.99 1899 0.198 -0.578 0.96 9 " a. estimated after Wetherall with pooled length-frequency distribution data; b. estimated with fixed L^ (after Wetherall); L, . asymptotic length (mm); K, growth constant (per y); t,,, age at zero length; n (for tagging-recapture data), number of growth increments; n (for age-length data), number of ages (= y) corresponding to a mean length, which was calculated from the growth ring data; r^ = 1 - RSS/TSS; RSS. residual sum of squares; TSS, total sum of squares. Population Dynamics of Bivalves 723 200 180 160 140 120 UJO 80 60 40 Loo K to V. antiqua 73.9 0.218 -0.161 T.dombeii 88.5 0.232 -0.658 E. macha 189.9 0.210 -0.578 8 0 2 4 6 age [yr] Figure 3. Growth curves of the VBCF and age-length data of three Chilean hivahe.s, \ . antiqua, T. domheii. and E. macha. VBGF pa- rameters (asymptotic length. L, |mm|; growth constant. K (per y|; age at zero length. t„) are given in the inset. Vertical hars indicate standard deviations of the mean lengths of the age-length data. year), and finally, T. domheii (7.8 and 26.7 g of AFDW/nr per vcar). The gonad production of the population was much higher than the somatic production. The following P/B ratios were cal- culated; T. doinbeii - 0.292. E. macha = 0.222. V. antiqua = 0.180. DISCUSSION Reproduction All three species of this study exhibited annual reproductive cycles with a short summer spawning period. The influence of latitudinal gradients on the reproductive strategies of G. solida. S. solida. and P. thaca arc discussed in Urban and Campos (1994). At 36°S. all three species have annual reproductive cycles with longer resting periods in winter, whereas at lower latitudes. G. solida from 14°S. Independence Bay. Peru (Ishiyama and Chavez 1990). S. solida from 30°S. Tongoy. Chile (Campos et al. 1993). and P. thaca from 23°S. San Jorge Bay. Chile (Henriquez et al. 1981). have biannual reproductive cycles, probably with shorter resting periods or continuous gonad activities throughout the year. Except for V. antiqua. no other examples of reproductive studies of the same species studied here were found in the literature. Tagelus divisus from 25°N. Biscayne Bay. FL (Fraser 1967). has a biannual body weight cycle with strong increases (followed by spawning events) from September to December and a second smaller increase from March to June. Ensis minor from 4\°N. Gulf of Manfredonia. Italy (Casavola et al. 1985). on the other hand, has a very short spawning period in March and April followed by a long resting period. All of these results confirm the well-known dependency of the reproductive strategy on latitudinal-dependent factors (e.g.. different temperature regimens): with increasing lat- itudes (in a northern or southern direction), reproductive strategies seem to change from continuous to biannual to annual cycles. There are. however, certain discrepancies, such as those of V. antiqua from the Bay of Ancud, Chile (Lozada and Bustos 1984), which has a continuous biannual reproductive cycle, although lo- cated further south (at 43°S) than the Bay of Dichato of this study. In(%> 1/At) 8 1 : ••■•\ 0 0° V. antiqua -1 p y= 10.04- 1.07 X >y^ Z= 1.084 0 ^v -3 T 1 1 1^' 1 3 5 7 9 II 13 2 "^X Sk,<^%/ >v T. dombeii 0 -o* > K -2 o ~ y = 5,78 - 0.83 x Z = 0.839 -4 Till o 1 1 1 1 0 2 4 6 8 10 12 14 T ooO°i y 0 - ''V cpo O o \« E. macha -2 ~ y =9*83- 1.08 X Z= 1.089 \ -4 Till 1 1 1 0 2 4 6 8 10 12 age [yr] Figure 4. Length converted catch curves of three Chilean hivalves, V. antiqua. T. dombeii. and E. macha based on pooled length-frequency samples (from April 1991 to March 1992). The VBGF parameters L„ K, and t,, required for this method are given in Figure 3. Solid sym- bols: used for calculations of total mortality (Z); open symbols: ex- cluded from calculations. Linear regression equation and estimated Z values are given. This could be because of local specific ambient conditions (e.g., currents, upwelling. and nutrients) of the Bay of Ancud that may make it a more productive region than the Bay of Dichato (which is a rather protected bay). Another reason could be that V. antiqua TABLE 4. Parameters of the SL (mm) AFDW (g) relationship for three Chilean bivalves, V, antiqua. T. dombeii, and E. macha.' Species a b r^ n V. antiqua T. dombeii E. machii -4.297 -5.277 -6.494 2.689 2.978 3.285 0.93 0.99 0.96 35 29 33 ' Log(AFDW) = = a -t- b Log(SL). 724 Urban -1 -1 g AFDW Individual yr IOt 3.0 2.5- ■ 2.0- ■ 1.5 :.o-- 0.5 ■• 0 20.0-1- (O) individual soniallc production (D) individual gonad produclion I I I I I I 10 20 30 40 50 60 70 XO 90 100 T. dombeii I I I 10 20 3(1 40 50 60 70 80 90 _ UK) E. macha 140 160 180 200 shell length Imm] Figure 5. Individual production of somatic tissues and individual go- nad production for different length classes of three Chilean bivalves, V. aniiqua, T. dombeii. and E. macha. is better adapted to the thermal regimen because it has a more southern distribution limit than C. solida. S. solida. and P. thaca (Urban 1994b). V. antiqiia is found up to the Magellan Strait, south ot Chile, 55°S (Urban and Tesch 1996). Growth There are no published growth rates for the species of this study, other than for V. iintiquu from the Bay of Yaldad, Chile 43°S (Clasing et al. 1994). Their initial VBGF parameters (L^ = 71.2 mm and K = 0.224/y) are almost identical to those of the population studied in Bay of Dichato (L.^ = 73.9 mm and K = 0.218/y). According to Knight (1968) and Theisen (1973). very misleading species-specific estimates of L„ can be obtained with data that does not cover most of a species growth range. Conse- quently. Clasing et al. (1994) did estimate a second set of VBGF parameters based on the maximum length of their population (L,,,^,^ = 80 mm) that resulted in K = 0. 183/y. Fortunately, the V. aniiqua population from Dichato had an estimated L.^ and an actual observed maximum length that were almost identical (L,^ = 73.9 mm. L,„^^ = 74 mm), which could be explained by com- mercial fishing activities. The VBGF parameters L^ and K are inversely related; conse- quently, it is not advisable to use only one of them, for example K. for growth comparisons. An index that considers the relationship between L, and K was developed by Munro and Pauly 1 1983; ' = Log(K) + 2 Log(L,_)|. Comparing the ' values (V. antiqiia. Dichato = 3.076; V. aiuiqua. Yaldad = 3.069; T. dombeii = 3.259; E. macha = 3.879) gives almost identical growth for the two V. aniiqua populations, a slightly higher value for T. dombeii, and the highest growth performance for E. macha. Mortality The mortality data of all six infaunal bivalve species from the Bay of Dichato are summarized in Table 6 (this study and Urban and Campos 1994). The information reveals some similar features. Maximum age ranges between 13 and 17 y; natural mortality. M, lies around 0.3 and is much lower than the fishing mortality, leading to high exploitation rates, E, around 0.7. Little comparative information exists in the literature. Saldivia (1981) estimated the mortality of V. aniiqua from the Bay of Ancud, Chile (43°S). He obtained Z = 0.43, M = 0.38, F = 0,05. and E = 0.12. As can be seen by the very low E value, in TABLE 5. Summary of production estimation for three Chilean bivalves, V. antiqua, T. dombeii, and E. macha. TABLE 6. Summary of mortality data for the six Chilean bivalve species obtained from Urban and Campos (1994) and from this study." Parameter Population gonad production (g AFDW/m- per y) Population somatic production (g AFDW/m- per y) Population mean biomass (g AFDW/m- per y) Population P/B ratio Mean abundance (g AFDW/m') Mean body weight (g AFDW) Species Species A„a, Z M F E T. dombeii E. macha V. antiqua G solida S. solida L3.5 12.7 0.846 0.916 0.291 0.310 0.555 0.606 0.656 0.662 270.3 22.9 98.5 P thaca 17.2 0.628 0.226 0.402 0.640 V. antiqiui^ 11.7 1.084 0.337 0.747 0.689 22.0 7.8 9.7 T. dombeii 14.7 0.839 0.266 0.573 0.683 E. macha 13.5 1.089 0.291 0.798 0.733 122.0 0.180 47.5 2.57 26.7 0.292 26.2 1.02 43.6 0.222 12 6 3.46 " Z is total mortality. M is mean natural mortality. F is fishing mortality, and E is exploitation rate. Also given is A„,^^ (y). the maximum age. "Calculated with L-, = 80.0 mm, taken from Clasing el al. (1994), because L^ found in this study is smaller than L^^,^. Population Dynamics of Bivalves 725 , (O) population somatic production 2 -1 N m "" (Q) population gonad production AFDW g m yr 20 T "-" xl40 140 IW) 180 2(X) shell length [mm] Figure 6. Distribution of somatic production and gonad production of the population as well as mean abundance for different length classes of three Chilean bivalves, V. antiqua, T. dombeii, and E. macha. 198 1 , this particular population must have been exploited at a very low level. Clasing et al. (1994), who studied the population dy- namics of V. antiqua from the Bay of Yaldad, reported Z = 0.66. M = 0.33, F = 0.33, and E = 0.50. which are in accordance with the data obtained in this study. According to Beddington and Cooke ( 1983), Caddy and Csirke (1983). and Francis (1974), an exploitation rate of E = 0.5 is above the optimal rate of exploitation for a fishery. Therefore, it can be assumed that all six bivalve species from the Bay of Di- chato arc severely overexploited. Indeed, fisherman from Dichato reported that catches in the Bay of Dichato have declined consid- erably during the last few years, so now. most of the fishermen exploit other banks or other resources. During the study period, little fishing activity was observed in the Bay of Dichato. The methods used here to calculate total and natural mortality are only valid under the assumption of a steady-state population. Unexploited populations under nonchanging biotic and abiotic conditions or populations exploited with the same exploitation rate for sonic time can be assumed to fulfill this assumption. The statements that catches have declined during the last years (with seemingly similar effort levels by commercial fishers) and that 'I o h-1 ^:-'i jh '127 24 "in I 21 12 5l» , ,■'5 Log(P/B) = -0.403 - 0.214 Log(MBW) r = 0.76, n = 64 -4-3-2-1 0 1 2 3 Log (MBW) [g AFDW] Figure 7. Comparison of somatic production of three Chilean bivalve species with other bivalve species of the same superfamilies (taken from the literature! by plotting the P/B ratio against the mean body weight (MBVV). Below, name of species, location of study area, and literature source. (I) same orders from other areas or related species from the same area — Va: V. antiqua, Td: T. dombeii, Em: E. macha. Bay of Dichato, Chile, 36°S (this study); Gs: G. solida, Ss: S. solida, Pt: P. thaca. Bay of Dichato. Chile, 36°S (Urban and Campos 1994): Gs90, Gs92, Gs93, Gs94: G. solida from four dilTerent years. Independence Bay, Peru, 14°S (Urban and Tarazona 1996); VaY: V. antiqua. Bay of Yaldad, Chile, 43°S (Clasing et al. 1994); Tdi: T. divisus, Florida (Fraser 1967); Es: Ensis siligua. South Wales, United Kingdom (UK) (Warwick et al. 1978): (III superfamily Tellinacea — I: Donax vittatus. South Wales, UK (Warwick et al. 1978): 2: Scrobicularia plana, Corn- wall. UK (Warwick and Price 1975); 3, 4: Abra alba, Kiel Bight, (iermany (Rainer 1985): 5-9: .4. alba, Manche, France (Dauvin 1986); 10: Abra nitida, Northumberland, UK (Buchanan and Warwick 1974); 11, 12: A. nitida. North Sea, Sweden (Josefson 1982); 13: Pharus le- gumen. South Wales, UK (Warwick et al. 1978); 14: Macoma ballhica. Nova Scotia, Canada (Burke and Mann 1974); 15, 16: M. balthica. North Sea, Denmark (Madsen and .Jensen 1987); 17: M. balthica, Cornwall, UK (Warwick and Price 1975); 18, 19: M. balthica. North Sea, Netherlands (Wolff and de Wolf 1977); 20, 21: Macoma calcarea, Arctic Sea, Greenland (Petersen 1978); 22-29: Tellinafabula, German Bight, Germany (Salzwedel 1980); 30, 31: T. fabula, Yorkshire, UK (Rees 1983); 32: T. fabula. South Wales, UK (Warwick et al. 1978); (111) superfamily Veneracea — 33-35: Callisia brevisiphonata. Sea of .lapan, Russia (Selin and Selina 1988); 36, 37: Chione cancellata, Flor- ida. USA (Moore and Lopez 1969); 38: Oosinia elegans, Florida, USA (Moore and Lopez 1970); 39—41: Mercenaria mercenaria, Southamp- ton (Hibbert 1976); 42. 43: M. mercenaria, Georgia. (Walker and Tenore 1984); 44, 45: Venerupis aurea, Southampton, UK (Hibbert 1976); 46: Venerupis decussata, Southampton, UK (Hibbert 1976); 47: Venerupis pullastra. North Sea, Norway (.lohannessen 1973); 48-50; Venus ovata. Manche, France (Dauvin 1985); 51: Venus striatula. South Wales, UK (Warwick et al. 1978). fishermen now exploit other resources (see previous paragraph) appear at first to suggest that the steady-state assumption is not valid anymore. However, the steady-state assumption may still be true for the following reasons; the exploited species from Dichato are rather long-lived individuals (13-17 y; Table 6); changes in fishing pressure/effort, on the other hand, are only a very recent (3—4 y) development (personal information of local fishermen). Therefore, it is deemed acceptable to use the standard methods for the estimation of Z, M, F, and E for the Dichato populations 726 Urban because they consisted mainly of old (and large) specimens (Fig. 6 in Urban and Campos 1994 and Fig. 6 in this study) when the study was conducted. Production The P/©B values of the six Chilean bivalve species from the Bay of Dichato (studied here and taken from Urban and Campos 1994) are plotted together with literature data on related bivalve species over their corresponding mean body weights (Fig. 7). The values of V. antiqua. T. dombeii, and E. macha are very similar to those of G. solida. S. solida. and P. tluica. No other production estimates exist in the literature except for V. imtuiiia from the Bay of Yaldad. Chile, 43°S (Clasing et al. 1994). and for G. solida from Independence Bay, Peru, I4°S (Urban and Tarazona 1996). Comparatively, the latter two populations have higher P/©B val- ues (Fig. 7). A possible explanation for the higher somatic pro- duction of the populations from the Bay of Yaldad and Indepen- dence Bay is that Clasing et al. (1994) and Urban and Tarazona (1996) described their investigation areas as being very produc- tive, with a high level of secondary production. Interestingly. Clasing et al. (1994) and Urban and Tarazona ( 1995) also reported strong fishing activities within their study areas. According to Urban and Campos (1994) SCUBA-diving fisherman select the larger bivalves in order to obtain better market prices. Thereby. the size-frequency distribution of the population is shifted toward smaller individuals with higher somatic production. Thus, if — because of recruitment — biomass remains constant, such popula- tions have higher P/B ratios (Hall et al. 1970). The P/B values of the Peruvian-Chilean populations seem to form a separate cluster with low productivity and high mean bio- mass values (Fig. 7). Length-frequency distributions show that all populations consist of old and large individuals (Fig. 6. this study; Fig. 6. Urban and Campos 1994). Hence, these individuals devote most of their surplus energy to gonad production rather than to somatic production (growth), leading to low P/B values. The rea- son is not clear; it could be. however, an adaptation to the high productivity of the upwelling system: the bivalves studied here are well adapted, such that they form dense populations all over their distribution range, dominating the benthic ecosystem (Urban 1994a). Most likely, recruitment to the population is space-habitat limited. ACKNOWLEDGMENTS The Dean of the Department of Oceanology. University of Concepcion. Chile, kindly permitted the use of facilities. Thanks are expressed to the scientific and technical staff of the University of Concepcion for friendly and helpful support. LITERATURE CITED Allen. K. R. 1971, Relation between production and biomass. J. Fislx. Res. Bd Can. 28:1537-1581, Beddington. J. R, & J, G. Cooke. 1983, The potential yield ol fi,sh stocks. FAO Fish. Tech. Pap. 242:1-47. Beverton. R. S. H. & S. J. Holt. 1956. A review of methods tor estimat- ing mortality rates in exploited fish populations, with special reference to sources on bias in calch sampling. Rapp P.-V. Reun. CIEM 140: 67-83, Buchanan. J. B. & R. M. Warwick. 1974. An estimation of benthic mac- rofaunal production in the offshore mud of the Northumberland Coast. J. Mar. Biol. Assoc. U.K. 54:197-222. Burke. M. & K. H. Mann. 1974. Productivity and production: biomass ratios of bivalve and gastropod populations in an Eastern Canadian Estuary. J. Fish. Res. Bd. Can. 31:167-177. Caddy. J. F. & J Csirke. 1983. Approximations to sustainable yields for exploited and unexploited stocks. Oceanof>r. Trop. 18:3-15. Campos. B.. R. Guerra & G. Bellolio. 1993. Bases biotecnoidgicas para la produccion masiva de juveniles de almejas de importancia econom- ica. Informe Final. Proyecto de Investigacion 502-91 . Fondo Nacional de Investigacion Cientifica y Tecnologica. CONICYT. Univcrsidad de Valparaiso. Vina del Mar. Casavola. N.. E. Rizzi. G. Marano & C. Saracino. 1985, Cicio nprodul- tivo e biometria di Ensis minor (Chenu) (Bivalvia: Solenidae) nel Golfo di Manfredoma. Oebaliu 11:439-449. Clasing. E.. T. Brcy. R. Stead, J. Navarro & G, Asenico. 1994. Popula- tion dynamics of Venus antiqua (Bivalvia: Veneracea) in the Bahi'a de Yaldad. Isla de Chiloe. Southern Chile. J. Exp. Mar. Biol. Ecol. 177:171-186. Crisp. D. J. 1984. Energy tlow measurements, pp. 284-372. In: N. A. Holme and A. D. Mclntyre (eds.l. Methods for the Study of Marine Benthos. Blackwell. London. Dauvin. J.-C. 1985. Dynaniique et production d'une population de Venus ovala (Mollusque-Bivalve) de la Bale de Morlaix (Manche Occiden- tale). J. Exp. Mar. Biol. Ecol. 91:109-123. Dauvin. J.-C. 1986. Biologie, dynamique et production dune population d'Abra alba (Wood) (Mollusque-Bivalve) de la baie de Morlaix (Manche Occidentale). J. Exp. Mar. Biol. Ecol. 97:151-180 Etmi. L. & T Brey 1994. Growth, productivity, and significance for fishery of the bivalve Egeria radiata (Donacidae) in the Cross River. Nigeria. Arch. Fish. Mar. Res. 42:63-75, Evcrsole. A, G 1989, Gametogenesis and spawning in North American clam populations: implications for culture, pp, 75-109, In: J, J. Mauzi and M. Costagna (eds.l. Clam Maricullure in North America. Dev. Aquacult. Fish. Sci. 19. Fabens A, J, 1965, Properties and fitting of the von Bertalanffy growth curve. Growth 29:265-285. Francis. C R. 1974. Relationship of fishing mortality to natural mortality at the level of maximum sustainable yield und the logistic stock pro- duction model. J. Fish. Res. Bd. Can. 31:1539-1542. Eraser, T. H, 1967, Contribution to the biology of Tai;elus divisus (Tell- inacea: Pelecypoda) in Biscayne Bay. Flonda. Bull. Mar. Sci. 17:1 1 1- 132. Hall. D. J., W, E, Cooper & E, E, Werner, 1970. An expenmental ap- proach to the production dynamics and structure of freshwater annual communities. Limnol. Oceanogr. 15:839-928. Henriquez, R., P. Barboza, E. Taipa & C. Toro. 1981. Variacion anual de la gonada de la almeja Protothaca thaca (Molina, 1782): analisis his- tologico. Resumen Jorn. Cienc. Mar 34. Hibbert. C, J, 1976, Biomass and production of a bivalve community on an inlertidal mud-Hat. J. E.xp. Mar. Biol. Ecol. 25:249-261, Hoenig, J, M, 1983, Empirical use of longevity data to estimate mortality rates, Fi.d,. Bull. 82:898-903. Ishiyama. C. V. & G. Chavez. 1990. Reproduccion de Gari solida (Ven- eroida, Psammobidae). Rev. Cieiuias U.N.M.S.M. 75:52-65. Johannessen, O. H. 1973. Length and weight relationship and the potential production of the bivalve Venerupis pullastra (Montagu) on a sheltered beach in Western Norway. Sarsia 53:41^8. Josefson, A. B. 1982. Regulation of population size, growth and produc- tion of a deposit-feeding bivalve: a long-term field study of three deep- water populations off the Swedish West Coast. J. Exp. Mar. Biol. Ecol. 59:125-150, Knight, W. 1968. Asymptotic growth: an example of nonsense disguised as mathematics. J. Fish. Res. Bd. Can. 25:1303-1307. Lozada, E. & H. Bustos. 1984. Madurez sexual y fecundidad de Venus Population Dynamics of Bivalves 727 anliqua (King & Broderip, 1835), en la bahia de Ancud (Mollusca: Bivalvia: Vcnendae). Rew Biol. Mar. Valparaiso 20:41-112, Madscn. P- B, & K. Jensen. 1987, Population dynamics of Mucwm; fcoWi- ica in the Danish Wadden Sea in an organically enriched area. Ophelia 27:197-208. Moore. H. B. & N. N. Lopez. 1969 The ecology of Chione cancellala. Bull. Mar. Sci. 19:131-148. Moore. H, B, & N, N, Lopez. 1970, A contribution to the ecology of the laniellibranch Dosinia elegans. Bull. Mar. Sci. 20:980-986, Munro. J. L. & D. Pauly. 1983. A simple method for comparing growth of fishes and invertebrates. Fishbyte 1:5-6. Pauly, D, 1979, Gill size and temperature as governing factors in fish growth: a generalization of von Bertalanffy's growth formula, Ber. Inst. Meereskde. Kiel. 63:1-156. Paulv. D, 1983, Length converted catch curves, A powerful tool for fish- eries research in the tropics, (Part 1). Fishbyle 1:9-13. Pauly. D. 1986. On improving operation and use of the ELEFAN programs. Part 11 Improving the estimation of L:^. Fishbyle 4:18- 20. Peterson, G H 1978. Life cycles and population dynamics of marine benthic bivalves from the Disko Bugt Area of West Greenland Ophelia 17:95-120. Press, W. H.. B. P. Flannery, S. A. Teukolsky & W. T. Vetterling. 1986. Numencal Recipes. The Art of Scientific Computing. Cam- bridge University Press, Cambridge. Rainer. S. F 1985. Population dynamics and production of the bivalve Abra alba and implications for fisheries production. Mar. Biol. 85: 253-262. Rees, H. L. 1983. Pollution investigations off the North-Last Coast of England: community structure, growth and production of benthic mac- rofauna. Mar. Envir. Res. 9:61-110. Saldivia, R. A. 1981. Estimacion de la mortalidad y otros factores en la pesqueria de almeja iVeiius annqua King y Broderip. 1831 ) en la Bahia de Ancud. Tesis para optar al li'tulo de Biologo Marino. Universidad de Concepcion, Concepcion. 80 pp. Salzwedel, H. 1980. Energy budgets for two populations of the bivalve Tellina fahula m the German Bight. VeroJ). Insl. Meeresforsch. Bremerh. 18:257-287. Selin. N. 1. & M. S. Selina. 1988. Production characteristics of the bi- valve mollusk Callisla brevisiphonala in Peter the Great Bay, Sea of Japan. Soviet J. Mar. Biol. 14:219-223. Urban. H.-J. 1992. Okologie likonomisch wichtiger Muschelarten im Auftriebssystem der Pazifikkiislc Lateinamerikas. Doctoral Thesis. University of Bremen. Bremen. 136 pp. Urban. H.-J. 1994a. On the adaptation of six infaunal bivalve species of Chile: differences in morphology, burrowing depth and substrate preference allow their coexistence. Arch. Fish. Mar. Res. 42:183- 193. Urban, H.-J. 1994b. Upper temperature tolerance of ten bivalve species off Peru and Chile related to El Nino. Mar. Ecol. Prog. Ser. 107:139- 145. Urban, H.-J. & B. Campos. 1994. Population dynamics of the bivalves Gari solida. Semele solida and Protothaca ihaca from a small bay in Chile at 36°S. Mar. Ecol. Prog. Ser. 115:93-102. Urban, H.-J. & J. Tarazona. 1996. Effects of El Nifio/Southem Oscillation on the population dynamics of a Gari solida population (Bivalvia, Psammobiidae) from Bahia Independencia, Peru (H'S). Mar. Biol. 125:725-734. Urban. H.-J. & C, Tesch. 1996 Aspects of the population dynamics of six bivalve species from Southern Chile, collected dunng the "Victor Hensen'-cruise to the Magellan Strait and the Beagle Channel in Oc- tober/November 1994, .Arch. Fish. Mar. Res. 44(3|, Theisen. B, F 1973, The growth o( Mxtilus edulis L. (Bivalvia) from Disko and Thule district. Greenland. Ophelia 12:59-77. von Bertalanffy. L. 1938. A quantitative theory of organic growth (inqui- ries on growth laws II). Hum. Biol. 10:181-213. Walker, R. L. & K. R. Tenore. 1984. The distribution and production of the hard clam, Mercenaria mercenaria. in Wassaw Sound. Georgia. Estuaries 7:19-27. Warwick, R. M., C. L. George & J. R. Davies. 1978. Annual macro- fauna production in a Venus community. Est. Coast. Mar. Sci. 7:215- 241, Warwick. R, M, & R, Price, 1975, Macrofauna production in an estuanne mud-flat, J Mar. Biol. Assoc. U.K. 55:1-18, Wetherall. J, A. 1986. A new method for estimating growth and mortality parameters from length-frequency data. Fishbyle 4:12-14. Wolff, W. J. & L. de Wolf. 1977. Biomass and production of zoobenthos in the Grevelingen Estuary, The Netherlands, Est. Coast. Mar. Sci. 5:1-24. Journal of Shellfish Research. Vol 15. No. 3. 729-733, 1996. SHELL LENGTH-MEAT WEIGHT RELATIONSHIPS OE OCEAN QUAHOG, ARCTICA ISLANDICA (LINNAEUS, 1767), FROM ICELANDIC WATERS GUDRUN G. THORARINSDOTTIR AND GARDAR JOHANNESSON Marine Rcxcmrh Institute P.O. Box 1390 SkiiUii>atii 4 121 Reykjavik, Iceland ABSTRACT Shell length-meat weight relationships were analyzed for the ocean quahog, Arctica islandica. for three geographic areas in Iceland. The allometric growth curves for the three areas were: northwest area, W = 0.0000567 L""'; north area, W = 0.000173 L" '^ east area, W = 0,0000929 L-^ "-. However, when the allometnc growth model was used, the three condition factors were not found to be significantly different from each other. Only one condition factor was therefore estimated for all areas but three different slopes: northwest area, W = 0.000637 L' "'; north area, W = 0.000637 L" ■"*; east area, W = 0.000637 L"". When assuming isometric growth for the three areas (i.e. , the slope = 3.0), the condition factors were found to be significantly different from each other; however, this model did not fit the data as well as the allometric one. The results from assuming either isometric or allometric growth indicated that the greatest relative meal weights for similar-sized quahogs were observed in the northwestern area. This may be due to the higher temperature and productivity of the northwestern area compared wilh the others. KEY WORDS: Arctica islandica. ocean quahog, shell length-meat weight relationship INTRODUCTION The ocean quahog, Arctica islandica. is one of the largest commercially fished bivalve molluscs inhabiting the marine waters of Europe and North America. It occurs in the North Atlantic along the East Coast of North America from Newfoundland to Cape Hatteras; on the coasts of Iceland, Faroes, the Shetlands. and the British Isles; and along the European Coast from the White and Barrents Seas to the Bay of Cadiz in Spain (Menill and Ropes 1969). It is found in waters as shallow as 4 m to as deep as 256 m, but the commercial fisheries occur on the continental shelf in waters from about 25 to 60 m deep (Merrill and Ropes 1969). Icelandic fisheries have only just begun, but investigations on distribution and abundance indicate a major resource that could support a commercial fishery (Thorarinsdottir and Einarsson 1996 in press). Much is known about the reproduction, growth, and age of the ocean quahog in North America (Rowell et al. 1990; Fritz 1991. Kraus et al. 1991, Kennish et al. 1994), and some recent infor- mation about recruitment and mortality rates also exists (Anony- mous 1993. Weinberg 1993. Anonymous 1995. Kennish and Lutz 1995). Aspects of ocean quahog density and distribution along the eastern coast of North America have been reviewed by Merrill and Ropes (1969. 1970). Fogarty (1981). Rowell and Chaisson (1983). and Chaisson and Rowell (1985). whereas the shell length-meat weight relationships of ocean quahog were examined by Murawski and Serchuk (1979), Murawski et al. (1982). and Fritz (1991). Little is known about populations of this species in the vicinity of Iceland. The objective of this study was to deter- mine the shell length-meat weight relationship of /\. islandica at the northwestern, northern, and eastern coasts of Iceland and to determine if there were significant differences in the mean weights calculated from direct observations and mean meat weights calcu- lated by using the length-weight relationship. MATERIALS AND METHODS Ocean quahog samples for the length-weight analysis were col- lected from Icelandic waters during assessment surveys conducted in January to March 1994 and May to June 1994 in the three principal areas, as shown in Figure 1 . A commercial hydraulic clam dredge was used for the sampling (1.5-m-wide cutting blade). The spacing between bars in the dredge was 34 mm. Sam- pling sites in each area were identified on sea charts according to the bottom topography and depths. Tows were then made from water depths ranging from 5 to 50 m. Individuals were randomly sampled for shell length measure- ments at each site and measured directly from dredge catches on the site (on-site measurements). The length-frequency distribu- tions for the three areas were compared by use of the Kolmogorov- Smirnoff two-sample test (see Thorarinsdottir and Einarsson 1996 in press!. Subsamples from each area were frozen and later thawed to determine shell length and meat weight relationships. Shell length was measured with vernier calipers to the nearest 0. 1 mm. and the wet meat weight was detennined to the nearest 0.1 g. The relationship between meat weight and shell length was assumed to be of the form: W = c LP (1) where W is the wet meat weight (in grams). L is the shell length (in millimeters), and c and p are coefficients to be estimated, i.e.. 24° 22° 20" 18° 16° 14° Figure 1. The locations of sample sites (black triangles) for A. island- ica, in northwest, north, and east sampling areas of Iceland, surveyed from January to June 1994, 729 730 Th6rarinsd6ttir and J6hannesson 18 16 14 12 10 B 6 4 2 0 - North ■ West n . 2265 7-75 4 ■ SD.14 4 ranqe = 17 - 118 Ill I,. 0> CT> CT) Ol CT> in in to 120 100 ■ 80 60 40 - 20 North ■ west - Allot netrtc yrowtti a nvo lsoi(»©trtc yfowtfi o ifvo 120 20 18 16 >> 14 0* 12 3 10 « it 8 # 6 4 2 0 North n = 1 1 00 T=74 5 SD = 11 5 range = 32- 107 -, ■■,l|l,l, ■ ■ 0)010)0)0)0)0001^0)0) • T-CNcnTintDr^cpo)OT- inioininioinioiomTT 100 ■ 80 60 40 ■ 20 0 ^ Allornetrti: ytowth curve _lso(ii©tilc growth ■"urve 20 40 60 80 100 ^0 ■ 20- East n= 1440 T=7S4 15- SD = 1 4 1 range = 17 108 ID- S' ,1 1 0 - .. ...111 . III... . jj^ojcn-vintDr^coo) inioiAioiAi^iAi^i^ T-CMCOTflOtOl^tOO) Shell length (mm) Figure 2. Shell length-frequency distribution for A. islandka (on-site measurements) from northwest, north, and east sampling areas of Iceland. Also shown are the mean shell length, standard deviation, and range. the condition factor and the slope. To estimate c and p. Eq. 1 was log transformed, making the estimation a matter of a simple linear regression: 8 100 Enst %o - Alor ieli1< yrowtri ciirve f> 80 ■ 60 Ison emc gfowtt-t ciiive o o o o 40 • o o 1 20 ^ 4 n ^ 0 J — 0 ^ 0.05) were Length- Weight Relationship in A. isuwdica 731 TABLE 1. The estimated shell length-meat weight relationship: type of relationship. In of the condition factor (Cond.l ± s.e. in brackets, the condition factor (Cond.), the slope ± s.e., r' total SSE (sum of squared errors), and p value from an F test indicating the significance of using an allometric relationship. Shell Length-Meat Weight Relationship Area Type In Cond. (±s.e.) Cond. Slope r2 n SSE p Value NW Isonietnc -9.45(0.014) 0.0790E-3 3.00 (Fixed) 0.966 277 14.20 Allometric -9.78(0.14) 0.0567E-3 3.08 (0.034) 0.967 13.91 0.018 N Isometric -9.74(0.020) 0.0588E-3 3.00 (Fixed) 0.773 134 7.17 — Allomeinc -8.66(0.55) 0.173E-3 2.75 (0.13) 0.779 6.87 0.054 E Isometric -9.61 (0.018) 0.0667E-3 3.00 (Fixed) 0.725 259 21.66 — Allometric -9.28(0.49) 0.0929E-3 2.92(0.11) 0.725 21.62 0.50 All Areas Isometric -9.57(0.011) 0.0698E-3 3.00 (Fixed) 0.914 670 51.64 — Allometric -9.11 (0.15) O.OlIOE-3 2.89 (0.34) 0.915 50.87 0.002 observed for the mean shell lengths from the northwestern (75.4 mm), northern (74.7 mm), and eastern (74.5 mm) areas. The shell length frequency distribution from the northwestern area was sig- nificantly different (p < 0.05) from the others (Thorarinsdcittir and Einarsson 1996 in press). An ANCOVA was carried out for the model: log(W,J = log(cJ -I- p^ log(L„ (3) Where W,, and L,j are the i-th measurements from area a, c^ and p^ are area-specific condition factors and slopes, and €,^ is a nor- mal error term (classic multiple regression). The results showed that the estimated slope (p) was highest in the northwestern area and the lowest in the northern area. In contrast, the condition factor (c) was highest in the north, followed by the east and then the northwest (Table I, Fig. 3). Reducing the model in Eq. 3 to a model that estimates only one condition factor and slope, log{c) + p log(L,^) was not significant: for log(c) + p log(L,^), the d.f. was 668; RSS (residual sum of squares) was 50.874; and R" was 0.9148. For log(cJ + p^ log(Li,J, the d.f. was 664; RSS was 42.405; R' was 0.9290; F value was 33.151; and p < 0.01. Similarly, reducing the full model in Eq. 3 to a model that estimates only one slope and three different condition factors, log(c^) -I- p log(L,^), was not significant (R" = 0.9283 and p value for differences in RSS was 0.036). On the other hand, re- ducing Eq, 3 to a model that estimates one condition factor and three different slopes, log(c) -I- p^ log(L,j), was significant (R^ = 0.9286 and p value for differences in RSS was 0. 1 39) (Fig. 4), For the special case of Eq. 3, when the slopes were fixed equal to 3.0 (isometric growth), the three condition factors were significantly different from each other: for log(c) -I- 3,0 Iog(L,J, the d.f. was 699 and the RSS was 51.637; for log(cJ -I- 3.0 log(L,J, the d.f. was 667; RSS was 42.926; F value was 67.679; and p < 0.01. However, the model with one condition factor and three ditlerent slopes was significantly better than the isometric growth inodel: for log(cJ -I- 3.0 log (L,J, the d,f, was 667 and RSS was 42,926; for log(c) -I- p^ log(L,J, the d.f. was 666; RSS was 42.658; F value was 4, 177; and p < 0.041 . The estimated slopes from the model with the same condition factor for all of the areas were 3.053, 2.981, and 3.010 for the northwest, north, and east areas, respectively, and they were all significant different from each other (Fig. 4). The results indicate that the quahog from the northwestern area generally contained more meat per unit shell length for the range of lengths considered than did individuals from the other areas (Figs. 3 and 4; Table 2). The mean meat weight of quahog in each area was estimated by using the length-weight relationship along with the observed length frequency distributions and compared with the mean weight from empirical data, as shown in Table 2. The two estimates were very similar, with the second method always within 2 s.e. from the first method (95% confidence inter- val). The average meat weights of quahogs from the northwestern, eastern, and northern areas were 38.5, 30,2, and 26.1 g, respec- tively. DISCUSSION In this study, the estimated length- weight relationships were significantly different among the three main areas. The meat weight for similar-sized quahogs were highest in the northwestern area, followed by the east; the lowest weight was observed in the northern area. No significant differences (p > 0.05) were observed among the average lengths of quahogs from the three areas, but there were significant differences among length-frequency distri- butions. Factors that are known to influence the relative condition of quahogs include physical and biologic variables such as water depth (Krausetal. 1991). temperature (Lutzet al, 1983), and food supply (Lutz et al. 1983, Grizzle and Lutz 1989, Kraus et al. 1991). Murawski and Serchuk (1979) calculated the length-weight re- 0 20 40 60 Shell length (mm) Figure 4. Estimated shell length-meat weight relationship for A. is- landica from northwest, north, and east sampling areas of Iceland. The model estimates the same condition factor for all three areas but three different slopes. 732 Th6rarinsd6ttir and J6hannesson TABLE 2. MMW of ocean quahogs from the three areas off Iceland and in the total area calculated by using (1) the observed meat weights in the subsample, (2) the shell length distribution and the estimated length-weight relationship, both from the subsample. and (3) the on-site length distribution with the estimated length-weight relationship from the on-site measurements. MMW (g) Meat Weight NW N Overall Average Observed Subsample Estimated On-Sitc 26.5 (s.e. = 1.1) 25.8 38.5 30.2 (s.e. = 1.6) 29.5 26 1 37.0 (s.e. = 1.2) 35 . 3 30.2 31.3 (s.e. = 0.74) 30.2 31.7 lationships in ocean quahogs from the Middle Atlantic shelf and confirmed allometric growth in two out of three areas. They ob- served that size-specific meat weights in the quahogs increased from north to south. This was attributed to be the stability of the thermal environment or related to density-dependent factors, but different stages in reproduction development could probably have affected their conclusion. Knowledge of both the physical and the biologic oceanography of inshore waters around the Icelandic Coast, unfortunately, is rather sparse to further explain the observed differences. Temper- ature data from 1908 to 1973 for stations within the three study areas show that the temperature range is 2.2-7.4°C at about 70-ni depth in the northwestern area. 1.5-6.7°C in the north area, and 1.2-6.9°C in the eastern area (Stefansson and Jonsdottir 1974). The primary production is also highest in the northwestern area (mean 184 g/cm" per year), followed by the eastern area (150 g/cni"^ per year) and the northern area (90 g/cm" per year), respec- tively (Thordardottir 1976). A more favorable thermal environment and higher primary pro- ductivity are important factors governing metabolic processes and partially explain the observed higher relative meat yields in qua- hogs from the northwest area. Murawski et al. (1982) observed small difference (4—11%) when comparing length- weight equa- tions from February and August for quahogs (65-115 mm) from the Middle Atlantic Bight. Winter samples were heavier in meat weight at a given shell length than summer samples. These dif- ferences were explained as related to weight changes associated with sexual development or statistical artifact. In ocean quahogs from New Jersey, size-specific somatic weight changed little through the year, which was suggested to be site differences in growth rate and reproductive cycling, and/or lack of synchrony of reproductive cycles of individuals at a given site (Fritz 1991). In this study, the samples fi"om the northwestern area were taken from January to March, whereas the samples from north and east were taken from May to June. Further studies are therefore nec- essary to determine the existence of seasonal changes in these length- weight relationships and the effects of sexual maturity. The weight of the soft tissue of the quahog from the northwest area increases from April until June and then falls in July and August, which is considered to be the main spawning time (Audunsson and Gunnarsson 1995). Inverse relationships have been observed between growth and intraspecific density in quahogs, which may be the result of com- petition for food or space (Eversole et al. 1990. Hurley and Walker 1994. Kraus et al. 1992. Rice et al. 1989). Beal and Kraus ( 1989) observed depressed growth in quahogs when the local density was over 600 individuals/m"^ (x = 49 mm); at lower densities (130- 323/m-). no difference in growth was observed. In this study, the mean density was only 25, 24, and 39 individuals/m" (x = 75 mm) in the northwest, north, and east areas, respectively (Tho- rarinsdottir and Einarsson 1996 in press); therefore, density- dependent factors have hardly affected the different growth curves observed in these areas. The length-specific meat weight in this study observed from the shell length-meat weight relationship was greater in all three areas than reported for quahogs from the eastern coast of the United States. In this study, calculated meat wet weight (MWW) for an individual of 95-mm shell length from the northwestern area sam- pled in winter was 70 g. and for the eastern and northern areas sampled in summer, it was 55.5 and 47.6 g. respectively. By use of the shell length-meat weight relationships reported by Muraw- ski and Serchuk (1979) for quahogs collected in winter off New Jersey. MWW of 36 g was calculated for an individual of 95-mm shell length. Murawski et al. ( 1982) reported MWW of 38 and 36 g in winter and summer samples, respectively, for 95-mm quahogs collected off New York, which is almost the same as the 41 and 39 g MWW calculated for quahogs of the same size collected at the same time of the year off New Jersey (Fritz 1991). ACKNOWLEDGMENTS We particularly thank Solmundur Einarsson, the leader of the scientific parties aboard the research vessels during the field sam- pling phases of the project. The manuscript was critically reviewed by Gunnar Stefansson and Karl Gunnarsson, to whom we owe our sincere thanks. LITERATURE CITED Anonymous. 1993. Report of the 15th Northeast Regional Stock Assess- ment Workshop (15th SAW). 1993. Stock Assessment Review Com- mittee (SARC) Consensus Summary of Assessments. NOAA/National Marine Fisheries Service. Northeast Fisheries Science Centre. Woods Hole. MA. 33 pp. Anonymous. 1995. Report of the 19th Northeast Regional Stock Assess- ment Workshop (19th SAW). 1995. Stock Assessment Review Com- mittee (SARC) Consensus Summary of Assessments. NOAA/National Manne Fisheries Service. Northeast Fisheries Science Centre. Woods Hole. MA. pp. 177-219. Audunsson, G. & E. Gunnarsson. 1995. Monitonng of algae toxins in ocean quahog iAntica islandica) from Onundartjordur. Fljolavik. and Adalvfk. NW-lceland, Apnl-November 1994. Rf Report 88. Icelandic Fish. Lab.. Reykjavik. Iceland, pp 1-18. Length-Weight Relationship in A. islandica 733 Beal. B. F. & M. G. Kraus. 1989. Effects of intraspecific dcnsUy on Ihe growth of Arclica islandica Linne inside field enclosures located in eastern Maine. USA. J. Shellfish Res. 8:462, Becker. R.. J. M. Chambers & A. Wilks. I98S The New S language. Wadsworlh. Belmont. CA Chaisson. D, R, & T, W. Rowell. 198.'i- Distribution, abundance, popu- lation structure, and meat yield of the ocean quahaug {Aniim islaiul- na) and Stimpson's surf clam {Spi sulci pi ihinnui) on the Scotian Shelf and Georges Bank. Can. hut. Rep. Fish. Aqua. Sci. 155:123 pp. Chambers, J. M. & T. J. Haslie. 1992. Statistical Models in S. Advanced Books and Software. Wadsworth and Brooks/Cole, New York. 608 pp Eversole. A G . J. G. Goddsell & P. J. Eldridge. 1990. Biomass. pro- duction and turnover of northern quahogs. Mercenaria mercenaria (Linnaeus. 1758). at different densities and tidal locations, J . Shellfish Res. 9:309- .M4, Fogarty. M. J, 1981. Distribution and relative abundance of the ocean quahog Arclica islandica in Rhode Island Sound and off Martha's Vineyard. Massachusetts. J. Shellfish Res. 1:33-39. Fritz, L, W. 1991. Seasonal condition change, morphometries, growth and sex ratio of the ocean quahog. Arclica islandica (Linnaeus. 1767) off New Jersey. U.S.A. J. Shellfish Res. 10:79-88. Grizzle. R. E & R. A. Lutz. 1989. A statistical model relating horizontal seston fluxes and bottom sediment characteristics to growth o( Merce- naria mercenaria. Mar. Biol. 102:95-105. Hurley. D H. & R, L Walker, 1994, Factors of bag mesh size, stocking density and quahog stocking size, which affect growth and survival of second year Mercenaria mercenaria (Linnaeus, 1758) in a coastal Georgia growout application. J. Shellfish Res. 13:303. Kennish. M. J. & R. A. Lulz. 1995. Assessment of the ocean quahog. Arclica islandica (Linnaeus, 1767). in the New Jersey Fishery, J. Shellfish Res. 14:45-52. Kennish. M. J., R. A. Lutz. J. A. Dobarro & L. W, Fritz. 1994, In-situ growth rates of the ocean quahog. Arclica i.\landica. in the Middle Atlantic Bight, J Sliellfish Res 13:47.3-478, Kraus. M G,. B, F, Beal. S. R, Chapman & L. McMartm. 1992. A comparison of growth rates in Arclica islandica (Linnaeus. 1767) be- tween field and laboratory populations, J. Shellfish Res. 1 1:289-294, Kraus. M, G,. B, F, Beal & L, McMartin, 1991 , Growth and survivorship of ocean quahogs. Arclica islandica (Linnaeus. 1767) in an intenidal mudflat in eastern Maine. J Shellfish Res. 10:290, Lutz. R. A.. J. G. Goodsell. M. Castagna & A. P. Stickney. 1983. Growth of experimentally cultured ocean quahogs (Arclica islandica L.) in north temperate embayments. J. World Mariciil. Soc. 14:185- 190, Merrill. A. S. & J. W. Ropes. 1969. The general distribution of the surf clam and ocean quahog. Proc. Nail. Shellfish Assoc. 59:40-45. Merrill. A. S, & J, W, Ropes, 1970, The distribution and density of ocean quahog. Arclica islandica. Am. Malacol. Union Bull. 36:19, Murawski. S. A.. J. W. Ropes & F. M. Serchuk. 1982, Growth of the ocean quahog, Arclica islandica. in the Middle Atlantic Bight. Fish. Bull. 80:21-34. Murawski, A. A. & F. M. Serchuk. 1979. Shell length-meat weight re- lationships of ocean quahogs. Arclica islandica. from the Middle At- lantic Shelf. Proc. Nail. Shellfish As.mc. 69:40-46. Rice. M, A,. C. Hickox & I, Zehra, 1989. Effects of intensive fishing effort on the population structure of quahogs. Mercenaria mercenaria (Linnaeus 1758) in Narragansctt Bay. J. Shellfish Res. 8:345-354, Rowell. T, W, & D, R, Chaisson. 1983, Distribution and abundance of Ihe ocean quahaug {Arclica islandica) and Stimpson's surf clam [Spisula polynyma) resource on the Scotian Shelf, Can. Ind. Rep. Fish. Aqua. Sci. 142:1-75. Rowell. T, W,. D, R, Chaisson & J, T, McLane, 1990, Size and age of sexual maturity and annual gamelogenic cycle in the ocean quahog, Arclica islandica (Linnaeus, 1767), from coastal waters in Nova Scotia, Canada, J Shellfish Res. 9:195-203, Stefansson, U, & S, Jcinsdottir, 1974, Near-bottom temperature around Iceland, Ril Fiskideildar 5, pp, 1-73 Thorarinsdottir, G, G, & S, T, Einarsson, 1996, Distribution, abundance, population structure and meat yield of the ocean quahog, Arclica is- landica (Linnaeus, 1767), in Icelandic waters. J. Mar. Biol. Assoc. U.K. (in press), Thordardottir. T. 1976, Preliminary assessment of the annual production in the shelf area around Iceland, ICES/CM1976/L:32, Planklon Comm. 4 PP Venables, W, N, & B, D, Ripley, 1994, Modem Applied Statistics with S-plus, Statistics and Computing, Springer- Verlag. New York. Weinberg. J, R, 1993. Ocean quahog populations from the Middle Atlan- tic to the Gulf of Maine in 1992, NOAA/NMFS. Northeast Fishenes Science Center Reference Document no. 93-02. Woods Hole, MA. 18 pp. Journal of Shellfish Resecmh. Vol, 15. No. 3. 735-740, 19%. TEMPORAL VARIATION AND TISSUE LOCALIZATION OF PARALYTIC SHELLFISH TOXINS IN THE NEW ZEALAND TUATUA (SURFCLAM), PAPHIES SUBTRIANGULATA LINCOLN MACKENZIE, Cawthron Institute Private Bag 2 Nelson, New Zealand DAVID WHITE, AND JANET ADAMSON ABSTRACT Changes in the paralytic shellfish poison (PSP)-to,\in profiles in populations of Tualua (Paphies suhnumgulata) inhabiting beaches in the Bay of Plenty were analyzed by high-performance liquid chromatography during the contamination phase caused by a bloom oi Alexandrium minulum in January 1993 and over a 6-mo penod 1 yr later, when low-level toxin residues persisted within these shellfish. During the peak of toxicity («412 ixg of saxitoxin [STX] equivalents/ 1 00 g). the toxin profiles consisted of various proportions of the carbamate (gonyautoxin) derivatives GTX,, GTX,, GTX,, GTXj. and neoSTX, and STX, with some traces of the decarbamoyl derivative dc-STX These profiles resembled those produced by the toxic dinoflagellate itself. One year later, when the toxicity had declined to a stable level of about 40 jj,g/100 g, only traces of derivatives other than STX remained and almost all of this toxin was sequestered within the siphons. The considerable length of time that toxin residues are retained, the tissue localization, and change with time in the spectrum of toxin derivatives in tuatua are very similar to those observed in other surfclam species elsewhere in the world. Analysis of toxin profiles in these shellfish provides a means of determining whether the observed PSP toxicity is the result of recent or long-pasl contamination episodes. KEY WORDS: Zealand Tuatua, surfclam. Paphies suhinanf>itUiki. paralytic shellfish poisoning, Alexaiuhnim mtnuium. Bay of Plenty, New INTRODUCTION During mid-summer (December to January) 1992-1993, the first documented case of paralytic shellfish poison (PSP) contam- ination of shellfish within New Zealand occurred when PSP toxins were detected by mouse bioassay in a variety of shellfish species from the Bay of Plenty. There was no human illness associated with the event, and the maximum recorded toxicity was a moder- ate 412 (i-g of STX equivalents/ 100 g of shellfish tlesh weight. It was quickly discovered (Chang et al. 1995) that the source of this contamination was a bloom of the toxic dinoflagellate Alexan- drium minutum apparently associated with a local upwelling event (Chang et al. 1996). As a result of this and other associated incidents (MacKenzie 1995, MacKenzie et al. 1995). a nation- wide shellfish biotoxin-monitoring programme, administered by the New Zealand Marine Biotoxin Management Board (N.Z.M.B.M.B.). was established in early 1993. This programme has involved the weekly sampling and testing of a variety of shell- fish species for aqueous and lipid soluble toxins from between 100 and 150 locations around the entire New Zealand coast. From September 1993 to June 1994, a weekly plankton-monitoring pro- gramme, incorporating the water column sampling of 17 sites along the north/east coast of North Island (including three sites in the Bay of Plenty), was also carried out. After the cessation of the 1993 A. minutum bloom, the PSP toxicity of most shellfish species within the Bay of Plenty declined rapidly. However, in tuatua. the PSP toxicity only gradually declined and it was not until 22 mo after the initial contamination event that this toxicity of these shell- fish became consistently undetectable by mouse bioassay. At the time of the peak in toxicity in January 1993. Bay of Plenty tuatua were analyzed by high-performance liquid chromatography (HPLC) to verify the results of the mouse bioassay s and to exam- ine the specific toxin composition. Between late 1993 and mid- 1994, when a stable base level of toxicity within these shellfish had been reached, a further series of analyses were done to com- pare with the original tests and to identify the reason for this continued contamination. The results of this investigation are pre- sented here. METHODS Tuatua were collected within the surf zone from Papamoa and Waihi beaches in the Bay of Plenty (Fig. 1 ) and transported live to the respective laboratories for mouse assays and HPLC analyses. Samples for mouse assays, carried out as part of the official marine biotoxin-monitoring programme were collected weekly from both of these beaches until August 1994. after which only Papamoa Figure 1. Map of the North Island, New Zealand, showing the loca- tion of shellfish-sampling sites in the Bay of Plenty. 735 736 Mackenzie et al. Beach was sampled. The mouse assays were performed at the Communicable Diseases Centre (Environmental Science Research Ltd.), Porirua, by use of the standard A.O.A.C. method (A.O.A.C. 1990). Between mid-January and mid-February 1993, the interim U.S.F.D.A. modification to the official A.O.A.C. extraction method (Hall 1991) involving the use of l.ON HCl was used. Thereafter, O.IN HCl was used according to the original standard method. Two samples (January 27 and 29, 1993) of tuatua collected from Papamoa Beach during the height of the bloom were ana- lyzed by HPLC, as were six samples (December 6, 1993; January 17, 1994; February 8, 1994; March 6, 1994; May 5, 1994; and July 4. 1994) from Waihi Beach. Extracts of whole shellfish and dissected body parts (from 10 shellfish) for HPLC analysis were prepared by boiling homogenized tissue in O.IN HCl (equivalent tissue weight/HCL volume), cooling, and adjusting the pH to 3. Extracts were cleaned up by passage through a "Sep Pak" (Wa- ters) C18 cartridge and by centrifugation through a 10.000 MW ultrafiltration membrane (Ultra Free C3-GC; Millipore); 10 (xL of this extract was injected into the HPLC system. The toxin spectra were resolved by use of a slight modification of the method de- scribed by Oshima et al. (1989b). These modifications involved the use of a GL Sciences Inc. Intersil C8 silica reversed-phase column and the following mobile phases for the different groups of toxin; (a) 2 mM tetrabutyl ammonium dihydrogen phosphate in acetate buffer (pH 5.8) for the C.-Cj toxins; (b) 2 mM heptane sulfonate in 10 mM ammonium phosphate buffer (pH 7.1) for gonyautoxins (GTX) 1-5; (c) 2 mM heptane sulfonate in 30 mM ammonium phosphate buffer (pH 7. l);acetonitrile, 100:5, for the saxitoxin (STX)-neosaxitoxin (neoSTX) group. The oxidizing agent was 7.0 mM periodic acid and 10 mM potassium phosphate buffer (pH 9.0). The fluorescent derivatives were measured on a Hitachi F-1000 fluorescence spectrophotometer at excitation and emission wavelengths of 330 and 390 nm, respectively. The stan- dards used for the calibration of the analysis were pure toxin standards prepared by the Laboratory of Food Hygiene (Depart- ment of Food Chemistry, Tohoku University, Sendai, Japan). The STX group standard mixture contained STX, neoSTX, and decar- bamoyl saxitoxin (dcSTX). The GTX group standard mixture con- tained GTX 1-5 (GTX1-GTX5), and the N-sulfocarbamoyl mix- ture contained toxins C1-C4. The calculation of the total toxicity of shellfish tissues (micrograms of STX equivalents per gram) from the HPLC data was accomplished with the conversion factors of Oshima (1995). On January 17, 1993, water samples at 3-m-depth intervals along three transects were sampled between Tauranga Harbour entrance and Waihi Beach. Between September 1 . 1993, and June 30, 1994, weekly phytoplankton samples from 3-m-depth intervals were collected from three sites within the Bay of Plenty — off Waihi Beach, Tauranga Harbour, entrance, and Ohiwa Harbour. Subsamples (10 mL) of Lugol's iodine-preserved samples were settled in Utermohl chambers and examined under an inverted microscope. All dinoflagellate species were documented and counted during whole-chamber scans. RESULTS The highest cell numbers of A . mumtum observed during the 1992-1993 bloom were 1.2 x 10"^ cells/L at a site off Tauranga Harbour on January 17, 1994. The peak in A. minutum abundance was coincident with the highest toxicity scores in tuatua from Bay of Plenty beaches. The A. minutum bloom in the Bay of Plenty declined soon after the maximum toxicity scores in shellfish within the bay were recorded, disappeared entirely within a few weeks, and was not observed within the phytoplankton in this area again between June 1993 and June 1994. A slow and somewhat erratic decline in toxicity within tuatua was observed after the disappearance of A. minutum from the plankton (Fig. 2), and shellfish toxicity did not consistently de- scend below the official action level of 80 (o-g of STX equivalents/ 100 g until December 1993, 10 months after the initial contami- nation event. Consistently negative mouse bioassay results from Papamoa Beach were not obtained until November 1994. Analysis of toxin profiles within whole-body extracts of tuatua (Fig. 3), collected from Papamoa Beach when toxicity was near its maximum (Fig. 3A), showed that STX and neoSTX (comprising 36 and 19% of total toxin body burden, respectively) were the major toxins, followed by various amounts of GTX1-GTX4. No N-sulfocarbamoyl derivatives were observed, although small amounts of the decarbamoyl derivative dcSTX were present. Pro- files from whole-shellfish extracts of six samples collected be- tween December 6, 1993. and July 3, 1994 (Fig. 3B), showed that Waihi Bch(D17) Papamoa Bch (D25) 1993 X 1- B 40U -»-WaihiBch(D17) 3sn -0- Papamoa Bch (D25) 300 200 150 100 i i i ___t_ i 50 r ^..^. -oOl3*Oc»« fn P ° D n n i s ^ ™ Apr Apr May Jun Jun ■Jul Aug ^ Aug Sep Sep 1 Ocl ' Mov 1 Dec 1 "^ - s s - " S ^ " s - "" S 2 "^ " S 01 c 1994 Figure 2. PSP toxicity of tuatua, determined by mouse bioassay, from Papamoa and Waihi Beaches, Bay of Plenty, during 1993 (A) and 1994 (B). The arrows indicate the dates on which samples of tuatua for HPLC-toxin analysis were collected. Toxins in P. subtriangulata 737 Figure 3. PSP-toxin pronies (mol%) in whole tuatua from Papamoa and Waihi beaches, 1993-1994. (A) Analysis of profiles in tuatua from Papamoa Beach collected on January 27 and 29, 1993 (n = 2). (B) Toxin composition in tuatua collected from Waihi Beach between De- cember 6, 1993, and July 3, 1994 (n = 6). Vertical lines indicate the range of values. STX was the overwhelmingly dominant derivative (94-100%). with only small traces (up to 7%) of GTXj present in some sam- ples. No N-sulfocarbamoyl derivatives were observed in whole- body extracts of these shellfish. Analysis of individual tissues withm the tuatua (Fig. 4) re- vealed the presence of trace amounts of other toxins (including some N-sulfocarbamoyl derivatives) in various parts of the shell- fish; however, the quantities of these and their contribution to the total toxicity of the shellfish were small in comparison to the predominance of STX residues within most tissues. Although only comprising a small proportion of the total body mass of the tuatua (1.3-2.5% by weight), the two siphons clearly contained the high- est concentrations of toxin residues (Fig. 5). The toxin was almost exclusively in the form of STX (Fig. 4). and high levels (up to 6.2 M-g of STX/g) were retained in these parts (Fig. 5A). Together the siphons contained, on average, 71% of the total toxin body bur- dens in these shellfish (Fig. 5B). DISCUSSION Since the initial discovery of PSP toxicity in the Bay of Plenty shellfish (Chang et al. 1995). the New Zealand shellfish biotoxin- monitoring programme has revealed several other instances of PSP-toxin contamination due to this dinoflagellate, the most im- portant of which occurred within the Marlborough Sounds in Jan- uary 1994 (MacKenzie 1994). With the exception of a recent event within the eastern Bay of Plenty that has been associated with AU'xandniim caienclla (MacKenzie unpublished), all cases of PSP-toxin contamination in New Zealand to date (where concur- rent phytoplankton data are available to substantiate this) have been attributable to blooms oi A. minutum. The absence of A. iminaum in the phytoplankton at the monitoring sites off Waihi Beach and Tauranga Harbour between September 1993 and June 1994 provides good evidence that the continued presence of PSP toxins within the tuatua in this region was not the result of con- tinuing cryptic contamination. The absence of PSP toxicity in other shellfish species (mussels, cockles, oysters) in this region over this period (M.B.M.B. records) provides further evidence that this was not the case. Because of the long retention time of PSP by tuatua, this spe- cies figures disproportionately in the New Zealand shellfish biotoxin contamination statistics. Between January 7. 1994, and January 6, 1995, 5,327 mouse bioassays for PSP were carried out (M.B.M.B. records) on a variety of shellfish species (of which 10% were tuatua) from an average of 102 sites per week distrib- uted around the entire coast of New Zealand. Only 2.9% of these bioassays returned positive results, of which 79% were in tuatua and all but one were the result of low-level residues in these shellfish from the Bay of Plenty and Northland regions. None of the PSP-toxin bioassay results in tuatua during this period ex- ceeded the 80 jxg of STX equivalents/ 100 g action level. 100 90 80 70 60 50 40 30 20 10 0 ■ Siphon (Inhalent) B Siphon (exhalent) D Hepatopancreas O Adductor a Foot D Remainder nail rtlilj II ,.n^i rirni n (9 (3 Figure 4. PSP-toxin profiles (mol%) in partitioned tissues of tuatua. (A) Samples collected on December 6, 1993. (B) Samples collected on June 4, 1994. 738 Mackenzie et al. — 4 « 2 A s i i (2 5) (14 1) (22 4) (20 1) (39 6) o £ a. m a. ■a 5 Figure 5. Mean toxin burdens in partitioned tissues of tuatua collected from Waihi Beach between December 6, 1993, and July 3, 1994 (n = 6). (A) Specific toxicity (jxg of STX equivalents/g) in various tissue types; the numbers in parentheses are the mean proportion ( % ) of each tissue type to total body mass. (B) The proportion that each tissue type contributed to total toxin body burden. Vertical lines indicate the range of values. The analysis of the PSP-toxin profile in an A. minuium isolate, established in culture from specimens collected during the 1993 Bay of Plenty bloom, was composed predominantly of neoSTX and STX, with lesser amounts of GTX,_4 and GTX,_, (Chang et al. 1996). The PSP-toxin profile produced by this isolate is rather different from the toxin profiles exhibited by A. minutum in other parts of the world. French isolates of A. minutum (Ledoux et al. 1993) have been shown to produce predominantly GTX-, and GTXj, with only trace amounts of STX, whereas Spanish isolates of A. minutum have been shown (Franco et al. 1994) to produce mainly GTX4 (80-90%), with lesser amounts of GTX, (10-15%) and only small amounts (3%) of GTX, and GTX^. Australian A. minutum (Oshima et al. 1989a, Franco et al. 1995) produces pre- dommantly GTX4 (77%) and GTXi (23%), with only small amounts (^10%) of GTX, and GTX3 and no evidence of any of the other 14 toxins known to occur naturally. The toxin profiles in tuatua collected at the time of the January bloom (Fig. 3A) were very similar to the toxin profiles observed in the dinoflagellate. The very small amounts of N-sulfocarbamoyl (C|_4) derivatives in the tuatua profiles do not provide unequivocal evidence that these toxins did not play a role in the contamination because samples were extracted with 0. IN HCl. It is likely that during this process, any N-sulfocarbamoyl derivatives that may have been present in these shellfish would have been hydrolyzed to their corresponding carbamate (GTX, 4) analogues (Cembella et al. 1993). The pres- ence of small amounts of dcSTX in the tuatua. but apparently not in the dinoflagellate (Chang pers. comm.), is consistent with a high capacity for in vivo PSP carbamate to decarbamoyl analogue conversion in surfclams (Cembella et al. 1993). The magnitude and nature of PSP toxin retention by shellfish vary depending on the shellfish species (Cembella et al. 1993, Cembella et al. 1994), and it has been known for many years (e.g., Medcof et al. 1947), and recently reaffirmed (Shumway et al. 1994). that surfclams are particularly prone to the retention of PSP toxins for considerable periods of time. The predominance of STX in the tuatua containing residual toxicity (Figs. 3 and 4) is typical of residues observed in other surfclam species elsewhere. Cem- bella and Shumway (1995) found that STX was the dominant toxin in most tissues of the surfclam SpisuUi solidissima in the Gulf of Maine, although there were substantial differences in the toxicity of different tissues as a proportion of total body toxin burden (digestive gland > mantle = gill > siphon = foot > adductor muscle). The butter clam (Saxidomus gigaiiteus) likewise is known to remain toxic for years after a contamination event (Quayle 1969) and to almost exclusively retain toxins in the si- phon, primarily in the form of STX. thus strongly resembling the nature of toxin retention by the New Zealand tuatua reported here. The predominance of STX as a long-term toxin residue within surfclams is due to either the accumulation and preferential reten- tion of STX over other derivatives acquired from the toxic di- noflagellates. or the conversion of other derivatives (GTXs) or precursors to STX within the shellfish tissues themselves. The experimental contamination of S. giganteus with a strain of A. catenella. which does not produce STX. suggests that the latter alternative is most likely in this species (Beitler and Liston 1990). Those investigators found that toxin derivatives rapidly accumu- lated throughout the shellfish tissues in proportions similar to those produced by the dinoflagellate; however, over a subsequent 58- day depuration period, the GTX derivatives declined as STX ac- cumulated within the siphon. Bricelj and Cembella ( 1995), on the other hand, found no evidence of the de novo appearance of STX in juvenile S. solidissima after experimental contamination with a strain of A. minutum rich in GTX. They hypothesized that the predominance of STX in wild S. solidissima populations was due to the exposure of these shellfish to STX-producing dinoflagel- lates. Because the strain of A. minuium responsible for the initial contamination of the Bay of Plenty tuatua does produce STX as a major part of its toxin profile (Chang et al. 1996), both mecha- nisms of STX residue retention are possible in these shellfish. The long-term retention of STX by tuatua is a nuisance for public health management because it only takes additional trace level contamination to raise the toxicity above the 80 |jLg of STX equivalents/ 100 g action level. On the other hand, where there is a suitable habitat for these shellfish, they do provide a means of mapping PSP contamination and the analysis of their toxin profiles gives an estimate of the frequency of PSP-toxin contamination events around the New Zealand coast. The results of the last 3.5 y of the N.Z. biotoxin-monitoring programme (January 1993 to June 1996) indicate that tuatua containing PSP-toxin residues are con- fined to the north/east coast of the North Island, although this species is routinely sampled on a few surf beaches elsewhere Toxins in P. subtriangulata 739 (North Island west coast and the South Island). This provides evidence for a low incidence of PSP causing blooms in these latter regions. Other New Zealand surfclam species (e.g.. the Pipi; Pii- phies ciiislnilis) may be equally good indicators for past intoxica- tion events, and some preliminary HPLC analyses (MacKenzie unpublished) have indicated the presence of trace amounts of PSP toxins in these shellfish at levels undetectable by the mouse bio- assay. All studies of the anatomical localization and retention of PSP toxin residues in surfclams have so far been carried out on northern hemisphere species, mainly from the northern North American continent. This investigation demonstrates that similar, although taxonomically and geographically very distant, bivalve molluscs occupying the same surf zone habitat in the southern hemisphere have similar characteristics in this regard and suggests that an ecological advantage may be gained by this capability. This lends weight to the theory (Kvitek 1993) that surfclams have evolved to use the retention of PSP toxins as a chemical defense mechanism against predation. Furthermore, the observations made here that toxin residues are sequestered in high concentration in the siphons support the hypothesis that this chemical defense strategy is spe- cifically aimed at protection against siphon-nipping fish. Tuatua are numerous on surf beaches throuahout New Zealand and. be- cause of their low intertidal zonation. are vulnerable to predation by sea birds, fish, and crustaceans and may provide a significant part of the diet of some of these species. If it is assumed that the ability to retain PSP toxins has evolved independently as a chem- ical defense device in indigenous surfclams. it leads to the con- clusion that although the PSP contamination of New Zealand shell- fish is a recently discovered phenomenon, it has in fact occurred for a long time and no doubt will continue to do so in the future. ACKNOWLEDGMENTS The authors are grateful to Associate Professor Yasukatsu Oshima. Department of Applied Biochemistry. Tohoku Univer- sity. Sendai. Japan, for providing training and the toxin standards used in the HPLC analyses. Thanks also to Eddie Ashcroft and John Tortell of East Bay Health. Tauranga. for the collection of shellfish and plankton samples; Wendy Gibbs for assistance in the preparation of the shellfish extracts: and the New Zealand Marine Biotoxin Management Board for the use of data from the shellfish biotoxin-monitoring programme. This research was funded under New Zealand Foundation for Science Research and Technology (FRST) contract CAW 601 with support (for the purchase of HPLC equipment) from the New Zealand Lottery Grants Board. LITERATURE CITED A.O.A.C. 1990. Official Method 959.08. pp. 881-882. In: K. Helnch (ed.). Paralytic Shellfish Poison biological method. 15th ed. Associa- tion of Official Analytical Chemists. Arhngton. VA. Beitler. M. K. & J. Listen. 1990. Uptake and tissue distribution of PSP toxins in Butter clams, pp. 257-262. In: E. Graneli, B. Sundstrom. L. Edler and D. M. Anderson (eds.). Toxic Marine Phytoplankton. Elsevier Science Publishers B. V.. New York. Bncelj. M. & A. Cembella. 1995. Fate of gonyautoxins in surfclams Spisula solidissima. grazing upon toxigenic Ale.uindrium. pp. 413- 418. In: P. Lassus. G. Arzul, E Erad, P. Genlien and C. Marcaillou- Le Baut (eds.l. Harmful Manne Algal Blooms, Lavoisier Intercept Ltd. Pans. Cembella. A. D. & S. E. Shumway. 1995. Anatomical and spatio- temporal variation in PSP toxin composition in natural populations of the surfclam Spisula soliilissima in the Gulf of Maine, pp. 42 1— 426. In: P. Lassus. G. Arzul, E. Erad. P. Gentien and C. Marcaillou-Le Baut (eds). Harmful Manne Algal Blooms. Lavoisier Intercept Ltd. Paris. Cembella. A. D., S. E. Shumway & R. Larocque. 1994. Sequestering and putative biotransformation of paralytic shellfish toxins by the sea scal- lop Placopecten magellaniciis: seasonal and spatial scales in natural populations. J. Exp. Mar. Biol. Ecol. 180:1-22. Cembella. A. D.. S, E. Shumway & N. I. Lewis. 1993. Anatomical dis- tribution and spatio-temporal variation in paralytic shellfish toxin com- position in two bivalve species from the Gulf of Maine. J. Shellfish Res. 12:389-403. Chang, F H,, D. M. Anderson, D. M. Kulis & D. Till. 1996b. Toxin production of Ale.xandriiim minulum (Dinophyceael from the Bay of Plenty, New Zealand. In: Proceedings of the Manne Biotoxin Science Workshop No 5. Coordinated by the New Zealand Manne Biotoxin Surveillance Unit, c;o MAF Regulatory Authority, Wellington, New Zealand. Chang. F. H.. L. MacKenzie. D. Till. D. Hannah & L. Rhodes. 1995. The first toxic shellfish outbreaks and the associated phytoplankton blooms in early 1993 in New Zealand, pp. 145-150. In: P. Lassus. G. Arzul. E. Erad. P. Gentien and C. Marcaillou-Le Baut (eds). Harmful Marine Algal Blooms. Lavoisier Intercept Ltd, Paris. Chang. F. H.. J. Sharpies & J. M. Grieve. 1996a. Temporal and spatial distribution of toxic dinofiagellales in Bay of Plenty New Zealand dunng the early 1993 toxic shellfish outbreaks, pp. 235-238 In: T. Yasumoto, Y. Oshima and Y. Fukugo (eds). Harmful and Toxic Algal Blooms. Intergovemmenlal Oceanographic Commission of UNESCO, Paris. Proceedings of the Seventh International Conference on Toxic Phytoplankton, July 12-16 1995, Sendai. Japan (in press). Franco. J. M.. P. Fernandez & B. Reguera. 1994. Toxin profiles of nat- ural populations and cultures of Ale.uindrium miniiliim Halim from Galician (Spain) waters, J. Appl. Phycol. b.lT^-TTi. Franco. J M., S. Fraga. M. Zapata. I Bravo. P, Fernandez & I. Ramilo. 1995. Comparison between different strains of genus Alexandrium of the minulum group, pp. 53-58. In: P. Lassus, G. Arzul. E. Erad. P. Genlien and C. Marcaillou-Le Baut (eds.). Harmful Marine Algal Blooms. Lavoisier Intercept Ltd. Paris. Hall. S. 1991. Interim revision of the A.O.A.C. PSP mouse bioassay protocol. U.S.F.D.A. memo 4/1/91. Washington. DC. Kvitek. R. G. 1993. Paralytic shellfish toxins as a chemical defence in the Butter Clam (Saxidomus giganteus). pp. 407-412. In: T. J. Smayda and Y. Shimizu (eds.). Toxic Phytoplankton Blooms in the Sea. Elsevier Science Publishers. B. V., Amsterdam. Ledoux, M.. M. Bardouil. J. M. Fremy. P. Lassus. I. Murail & M. Bohec . 1 993 . Use of HPLC for toxin analysis of shellfish contaminated by an Alexandrium minulum strain, pp. 413—418. In: T. J. Smayda and Y. Shimizu (eds). Toxic Phytoplankton Blooms in the Sea. Elsevier Science Publishers. B. V.. Amsterdam. MacKenzie. L. 1994. A bloom oi Alexandrium minulum in Anakoha Bay, Marlborough Sounds (Dec 1993-Jan 1994), Seafood N. Z 2:50-52. MacKenzie, L. 1995. Alcxandrnim and PSP in New Zealand: another chapter in the marine biotoxin saga. Seafood N. Z. 2:72-75. MacKenzie, L.. L. Rhodes, D. Till, F. H. Chang, H. Kaspar, A. Hay- wood, J. Kapa & B. Walker. 1995. A Gymnodinium sp. bloom and the contamination of shellfish with lipid soluble toxins in New Zealand Jan-April 1993. pp. 795-800. In: P. Lassus, G. Arzul, E. Erad, P. Gentien and C. Marcaillou-Le Baut (eds.). Harmful Marine Algal Blooms. Lavoisier Intercept Ltd., Paris. MedcoL J. C, A. H. Leim, A. B. Needier. A, W, H. Needier, J. Gib- 740 Mackenzie et al. bard & J. Naubert. 1947. Paralytic shellfish poisoning on the Canadian Hashimoto and Y. Ueno (eds.). Mycotoxins and Phycotoxins '88. Atlantic coast. Bull. Fish. Res. Board Can. 75:1-32. Elsevier Science Publishers. New York. Oshima, Y. 1995. Postcolumn derivatisation liquid chromatographic ., . , ... ■ ,1^ , . , . ^ < ^ , ^o C-.0 C1-, Quayle, D. 1969. Paralytic shellfish poisoning in British Columbia. Fish. method for paralytic shellfish toxins. y./l.O.A.C. //». 78:528-532. pnv<" d ii' a o Oshima. Y., M. Hirota, T. Yasumoto, G. M Hallegraeff. S. I. Blackburn & D. A. Steffensen. 1989a. Production of paralytic shellfish toxins by Shumway. S. E.. S. A. Sherman. A. D. Cembella & R. Selvin. 1994. the dinoflagellate /IfevanrfriHm »»«»/»»; Halim from Australia. Mp/Jon Accumulation of paralytic shellfish toxins by surfclams. Spisula so- Siiisan Gakkaishi 55:925. lidissima (Dillwyn. 1897) in the Gulf of Maine: seasonal changes, Oshima. Y.. K. Sugino & T. Yasumoto. 1989b. Latest advance in HPLC distnbution between tissues, and notes on feeding habitats. Natural analysis of paralytic shellfish toxins, pp. 319-326. In: S. Natori, K. To.xins 2:236-251. 1 Joiirmil of Shellfish Research. Vol. 15. No. 3. 741-745. 1996. REPRODUCTIVE CYCLE OF LAEVICARDWM ELATUM (SOWERBY, 1833) (BIVALVIA: CARDIIDAE) IN BAHIA CONCEPCION, BAJA CALIFORNIA SUR, MEXICO MARCIAL VILLALEJO-FUERTE, BERTHA PATRICIA CEBALLOS-VAZQUEZ, AND FEDERICO GARCIA-DOMINGUEZ Centra Interdisciplinario de Ciencias Marinas Instituto Politecnko Nacional Apdo. Postal 592 La Pa:. B.C.S. 23000. Mexico ABSTRACT The reproductive cycle of the cockle Laevicardium elaium in Bahia Concepcion. B.C.S. Me.xico, was studied from October 1988 to October 1990. Microscopic analysis established that L. elaium is a functional hermaphrodite with male and female follicles intermingled in the gonad. The development and spawning of male and female gametes were synchronous. Spawning occurred between October and April (18-23°C) and was related to a high food availability and to a lower condition of the organisms KEY WORDS: reproductive cycle, bivalves, Luevicaniium. histology INTRODUCTION The cockle Laevicardiuin elatum (Sowerby 1833) is distributed from southern California to Panama and lives in sandy bottoms; it has a length to 150 mm and is the largest of the living species of the family Cardiidae (Keen 1971). This species is considered a potential fishery in Baja California Sur (B.C.S.) (Baqueiro et al. 1982). Argopecten circularis and Megapitaria squalida are the principal species of economic importance of Bahi'a Concepcion. When the stocks of A. circularis and M . squalida are depleted, L. elaium is harvested and has a good acceptance because of its white meat. It is mostly harvested for its shell for ornamental purposes. Despite its biologic and economical importance, there are no pub- lished records on the biology of this species from Bahia Concep- cion waters. In contrast, Cardium edule. Cardnim gtaucum. and Cardium hauiuense have been well studied in Europe (Boyden 1971 . Kingston 1974, and Wolowicz 1987). There are also a small number of studies on the reproductive biology of some Cardiidae species, e.g.. Clinocardium nuttcdlii at Isla San Juan. United States, and Laevicardiuin laevigaluni from Venezuela (Gallucci and Gallucci 1982, Penchaszadeh and Salaya 1983). This study was carried out to describe the reproductive cycle and the spawn- ing season of Z,. elaium in Bahia Concepcion. B.C.S.. Mexico and its relation to the temperature, food availability, and condition (fatness) of the organisms. MATERIALS AND METHODS Between October 1988 and October 1990. in Bahia Concep- cion. B.C.S. (26°55'-26°30' N and 112°-111°40' W) (Fig. 1), 30-40 adult specimens of L. elaium were collected by skin diving between 2- and 6-m depth. The surface water temperature was recorded at the time of sampling. Shell heights and total and soft body wet weights were recorded for each clam. The clams were fixed in a neutral formalin solution. Tissue sections were taken at a standard point halfway between the di- gestive diverticula and the foot. These tissue sections were dehy- drated in alcohol and embedded in paraffin. Section (7jjLm) were placed on slides and stained with hematoxylin-eosin (Humason 1979). A modification of the developmental stages established by Gallucci and Gallucci (1982) for C. nuttallii was used to catego- rize follicles. N -9^ / \ / 26"55' / 1 \ \ (\ 1 ^—"^ 1 26 -ao- \ ks-v^ --- — ^ r i ^ \ PACIFIC \ U ^ MEXICO OCEAN \ \ \ Figure 1. Location of Bahia Concepcion, B.C.S., Mexico. ^Sampling area. Indifferent Stage This stage is characterized by a total absence of gametes: there- fore, it is not possible to distinguish the sex. The connective tissue occupies almost all of the space (Fig. 2A). Developing Stage In the female, there were rounded oocytes along with oocytes attached to the follicle wall. Some detached oocytes occurred. In 741 742 ViLLALEJO-FUERTE ET AL. '^ J Figure 2. Photomicrographs of gonadal stages of L. elalum. (At Indifferent. (B) Developing. (C) Ripe. (D) Partially spawned. (E) Spent. Bar = 55 nm. the male, varying quantities of spermatogenic cells were present follicles filled by spermatozoa arranged in characteristic bands (Fig. 2B). (Fig. 2C). Ripe Stage Partially Spawned In the female, most oocytes were free within the follicles, but In the female, large spaces mside the follicles and between free some oocytes remained attached to the follicle wall. In the male, oocytes were present. Some follicles were completely devoid of Reproductive Cycle of L. elatvm 743 oocytes. In the male, a marked decrease in the quantity of sper- matozoa was observed. Large spaces inside the folhcles occurred. In some follicles, only a few residual spermatozoa were present (Fig. 2D). Spent At this stage, some unspawned oocytes and spermatozoa were observed within follicles. The gametes were being phagocyted by amebocytes (Fig. 2E). The diameter of at least 100 oocytes, in each of six randomly selected females, was measured with an eyepiece graticule calibrated with a stage micrometer. The mea- surements were made along the longest axis of the oocyte, sec- tioned through the nucleus. From these data, mean oocyte size was obtained. Individuals with few measurable oocytes and extensive phagocytosis were not considered, following the criteria of Grant and Tyler (198.^a and b). The condition was estimated by the use of Fulton's equation (Hile 1936): W where CF is condition factor. W is soft body wet weight (in grams), and H is shell height (in millimeters). RESULTS Clams ranged in size from 65 to 155 mm in shell height, with the mode at 125 mm in height (96% of the organisms fell between 95 and 145 mm in height). L. elatutn is a functional hermaphro- dite. In the male and female acinis, the gametes were in the same developmental stage. Male and female follicles were intermingled in the gonad; therefore, the gonad did not have well-differentiated male and female glandular areas. The reproductive cycle of L. elatum from Bahia Concepcion. B.C.S., is summarized in Figure 3. From June to August, most cockles were inactive, as determined by the presence of indifferent and spent stages. The developing stage was minimal in this period The ripe stage was present from October to April 1988 and from >- O z UJ 13 o OCT FEB JUN OCT MAR AGO NOV ABR AGO ENE JUN OCT 1988 1989 1990 I I INDIFFERENT rZ2 DEVELOPMENT ^J RIPE ^1 PART SPAWN [H]U SPENT Figure 3, Reproductive cycle oi L. elatum. Relative frequency of go- nadal stages between October 1988 and October 1990. January to March and October 1990. The partially spawned stage, which is the indicator of the period of spawning, was present from February to April and October 1989 and January to March and October 1990. The measurements of oocyte diameters (Fig. 4) shows that gametogenesis and maturity are rapid processes. Oocyte develop- ment was faster from October to November 1988, when the high- est frequencies of mature individuals occurred. The oocytes were fully developed (56.4 \i.m, mean diameter; s.d. = 9.9) from Feb- ruary to April 1989 and January to March 1990, when the highest frequencies of partially spawned stage were present. The higher values of condition factor were found in April and August 1989 and in October 1990, whereas the lower values of condition factor occurred in November 1988, October 1989, and June 1990. There were two periods of recovery — November to April 1988 and June to October 1990 (Fig. 4). The minimum water O < Q Z o o E CC < D LU H > O o o a: Z) H < IT LU \- o m o) I E I z CO <"z 0£ II Q- CL OCT FEB NOV APR 1988 JUN OCT MAR AUG AUG JAN JUN OCT 1989 1990 Figure 4. Mean cundiliun factor values of L. elatum of Bahia Concep- cion, mean oocyte diameters of female, water temperature, and pho- tosynthetic pigment concentration. 744 VILLALEJO-FUERTE ET AL. temperature (18°C) was recorded in February, and the maximum (30°C) was recorded in August (Fig. 4). The photosynthetic pigment concentration mg of chlorophyll per cubic meter) from Bahia Concepcion. B.C.S.. obtained from satellite-derived information (Tran et al. 1993). was greater in cold months than in warmer months. There was a marked winter bloom that reached the maximum value in January (3.92 mgCl/m^). The minimum value was found in August (0.22 mgCl/m^) (Fig. 4). DISCUSSION The characteristics of gametogenesis in L. elation were similar to those described forC. nutlallii (Gallucci and Gallucci 1982). In both species, male and female follicles were shown to develop in phase with each other and the gametes of both sexes were spawned at about the same time. Among the majority of the hermaphroditic bivalve species, the condition of having intermingled male and female follicles is anomalous and they are reported as unusual cases (Penchaszadeh and Salaya 1983). Nevertheless, other func- tional hermaphrodite bivalve species, such as C. nutlallii. exist (Gallucci and Gallucci 1982). In Bahi'a Concepcion, the only other hermaphrodite bivalve species that has been studied is A. circii- laris (Villalejo-Fuerte and Ochoa-Baez 1993). The mean diameter of oocytes of L. elatum is similar to that of other bivalves found in Bahia Concepcion, A. circidaris and M. squalida (Villalejo-Fuerte and Ochoa-Baez 1993, Villalejo-Fuerte et al. unpublished data). Comparmg oocyte diameters with go- nadal stages in L. elatum, it is clear that minimum diameters coincide with the developing stage and maximum diameters coin- cide with the mature and partially spawned stages. Therefore, the oocyte diameters are reflective of the gametogenic cycle. Simi- larly, in Mercenaria spp. from Florida and in Glycymeris gigantea from Bahia Concepcion, maximum oocyte diameters were ob- served in conjunction with the period of maturation and spawning (Hesselman et al. 1989, Villalejo-Fuerte et al. 1995). Temperature is an important environmental factor in the regu- lation of bivalve reproduction (Sastry 1979). Differences in the timing of gametogenesis and spawning within a sjjecies over a latitudinal range occur because critical temperatures are attained at different times (Hesselman et al. 1989). The reproductive cycle of L. elatum in Bahia Concepcion shows a clear seasonality related to the water temperature. The inactive period occurs from June to August, with water temperatures of 28-30°C. Gametogenesis and spawnmg occur from October to April, starting when temperatures are declining and continuing during the cooler months ( I8-23°C). A similar relationship between the temperature and gonadal activ- ity has been observed for A. circularis (Villalejo-Fuerte and Ochoa-Baez 1993), G. gigantea (Villalejo-Fuerte et al. 1995) in Bahia Concepcion, and Mercenaria spp. in Florida (Hesselman et al. 1989). Higher temperature inhibits gametogenesis. Food availability has been related to the timing of bivalve reproduction (Sastry 1979, Bayne and Newell 1983, MacDonald and Thompson 1985, Jaramillo et al. 1993). The spawning in bivalves might be synchronized to coincide with maximum food availability for larval development (Seed 1976). Martinez-Lopez and Garate-Lizarraga (1994) found high concentrations of nanno- plankton in Bahia Concepcion in February; this represents a source of food for larvae oi A. circularis and other mollusks of commer- cial importance. The reproductive cycle of L. elatum has a seasonality related to food availability, expressed as the concentration of photosynthetic pigments. This relation is not observed in other bivalves like Hin- nites giganteus (Malachowski 1988). The spawning season of L. elatum coincides with the highest food availability, giving larvae the chance of exploiting the winter phytoplankton bloom. In this context, L. elatum is a conservative species under the scheme of Bayne ( 1976). who classified the reproductive strategy of bivalves according to the relation between the spawning and the storage cycles. The fluctuations of the condition factor are associated with the reproductive or nutritional condition of the mollusks (Searcy- Bemal 1984). In L. elatum. the maximum values of condition factor found in the inactive period (April and August 1989 and October 1990) could be produced by the accumulation of reserve substances during this period, which will be used during ripening. This has been observed for other bivalves (Sastry 1979). However, the fluctuations of the condition factor also may be a consequence of the water content in the soft body or changes in the mass of nutritive tissue (Giese and Pearse 1974). Then, the condition fac- tor is not a reliable indirect indicator of the spawning season and it is necessary to examine gonads microscopically. ACKNOWLEDGMENTS We are grateful to Direccion de Estudios de Posgrado e Inves- tigacion del Instituto Politecnico Nacional (IPN) for founding this work and to Comision de Operacion y Fomento de Actividades Academicas del IPN for the grants to M. Villalejo-Fuerte and F. Garcia-Dominguez. Thanks to Consuelo Gonzalez O. for her help with editing the English manuscript. LITERATURE CITED Baqueiro, C. E., J. A. Masso & H. Guajardo. 1982. Distribucion y abun- dancia de moluscos de importancia comercial en Baja California Sur. Secretaria de Pesca. Instituto Nacional de la Pesca, Mexico. 32 pp. Bayne. B. L. 1976. Aspects of reproduction in bivalve molluscs, pp 432-448. In: M. Wiley (ed.). Estuarine Processes, vol. 1. Academic Press, New York. Bayne. B. L. & R. C. Newell. 1983. Physiological energetics of marine molluscs, pp. 491-498. In: A. S. M. Saleuddm & K. M. Wilbur (eds.). The Mollusca. vol. 4. Academic Press, New York. Boyden. C. R. 1971. A comparative study of the reproductive cycles of the cockles Cerastoderma edule and C glauciim. J . Mar- Biol. Assoc. U.K. 51:605-622. Gallucci V. F. & B. B. Gallucci. 1982. Reproduction and ecology of the hermaphroditic cockle CUnocardium nutlalli (Bivalvia: Cardiidae) in Garrison Bay. Mar. Ecol. Prog. Ser. 7:137-145. Giese. A. C. & J. S. Pearse. 1974. Introduction: General principles, pp. 1—49. In: A. C. Giese & J. S. Pearse (eds.). Reproduction of Marine Invertebrates, vol. I. Academic Press, New York. Grant. A. & P. A. Tyler. 1983a. The analysis data in studies of inverte- brate reproduction. I. Introduction and statistical analysis of gonad indices and maturity indices. Int. J. Invert. Reprod. 6:259-269. Grant. A. & P. A. Tyler. 1983b. The analysis data in studies of inverte- brate reproduction. II. The analysis of oocyte size/frequency data, and comparison of different types of data. Int. J . Invert. Reprod. 6:271- 283. Hesselman. D. M.. B. J. Barber & N J. Blake. 1989. The reproductive cycle of adult hard clams. Mercenaria spp. in the Indian River Lagoon. Flonda. J. Shellfish /fcs. 8:43^9. Hile. R. 1936. Age and gwwtb of Cisco Leucichtys arledi (Le Sueur) in the lakes of the northeastern Highlands. Wisconsin. Bull. U. S. Bureau Fish. 48:209-317. Reproductive Cycle of L. elatum 745 Humason. G. L. 1979. Animal Tissue Techniques. W H. Freeman and Co., San Francisco. 661 pp. Jaramillo. R.. J. Winter. J. Valencia & A. Rivera. 1993. Ganietogenic cycle of Ihechiloe Scallop (C/i/11 mm. Data were tabulated separately on each of these size classes, and means were tested for significance by a f-test. Mussel mortality resulting from their handling during size- selection procedures was also compared. Immediately after pick- ing or sieving, the mussels were placed in beakers and stored in 80-L aquaria equipped with filters and air stones and filled with unchlorinated tap water (4°C). To retain mussels, the mouth of each beaker was covered with nylon screening held in place by a rubber band. Beakers were stacked on their sides in the aquaria, with 3=10 cm between beaker mouths and aquarium walls for water circulation. After 7 days, mortality was scored. Gaping mussels or those that did not close their shells when gently pried apart were considered dead. Picking Procedure Three timed trials (185, 231 , and 378 min) were conducted. In each, mussels were detached from rocks by grasping their byssal threads with fine-tipped forceps and gently pulling the byssus from the substrate. A gauge was used for rapidly separating the de- tached mussels into the three size classes (Fig. 3). When placed in this gauge, if a mussel's length exceeded Slot A, it was counted in the >ll-mm size class; if less than or equal to Slot B, it was 747 748 BiSS ET AL. ■^ 'M HIH Hm p^Hsp V^H 1.. ^it^'ji^^^^^l - „., ^-^iijl 1 JB ^1 ^ ■n ^^3 Figure 1. Zebra mussel passing downward tlirough a sieve pore lengthwise. Figure 2. Zebra mussel being selected in a pore by its largest-cross-sectional diameter, i.e., height. counted in the <6-mm size class. All other mussels were counted in the 6- to 1 1 -mm size class. After this size separation, mussels of the same size class were placed together in beakers and held as outlined above. Sieving Procedure Preliminary trials suggested that sieves with 5.6- and 2.8-mm pores sizes would retain mussels 6-1 1 mm in length between them. Timed sieving trials (46, 60, and 186 min) were thus con- ducted with this pair of sieves to test the accuracy of their size selection. With a chisel-tipped Exacto* knife, care was taken to cut the byssal threads and not to pry the mussels off of rocks. Mussels were separated from each other either by gently rubbing them between the fingers or by using the knife to cut byssal threads. They were then layered in the top sieve (i.e., 5.6 mm) to a maximum thickness of ca. 1.5 cm; a thicker layer risked ob- structing free passage of individual mussels through pores. For convenience, a third sieve with a 1 .0-mm pore size was used to trap mussels passing through both sieves. A sieve cover was placed on top of the three sieves, and a 2-cm-wide elastic band was attached to keep the cover and sieves firmly together. Sieving was performed in a 18-L pail filled with unchlorinated tap water. The sieves were submerged into the pail of water rapidly, thus causing the mussels to be flushed upward into the water column. When air bubbles stopped emerging from around the sieve cover (ca. 10-15 sec), the sieves were lifted completely out of the water. At this point, the mussels were floating in the water column, and as the water drained out, they were drawn lengthwise down through the pores. This submerging process was repeated four times. The mus- sels still remaining on the 5.6-mm sieve (presumably >11 mm long), those collected between the 5.6- and 2.8-mm sieves (pre- sumably 6-1 1 mm long), and those that had passed through both sieves (presumably <6 mm long) were transferred into separate beakers and stored as described above. After the 7-day mortality check, the length of each mussel was measured with the gauge (Fig. 3) to check the accuracy of the size class separation. Additional sieves were used in pairs to examine the size range of zebra mussels for which they would select. The pore sizes of these sieves were: 1 .00, 2.00, 2.36, 2.80, 3.35, 4.00, 4.75, 5.60, 6.30, 9.50, and 12.50 mm. The zebra mussels from the Mohawk River that were used in these experiments ranged from 2. 1 to 31.2 mm in length. RESULTS AND DISCUSSION Regression analysis revealed a strong correlation between shell length and height of D. polymorpha from the Mohawk River (height = 0.65 length" ""; r^ = 0.99); for any given length, only a small range existed in height, particularly with smaller mussels (Fig. 4). Although the sieving procedure could be used to separate size classes of any zebra mussel population, the results ot our sieving trials are only applicable to mussel populations whose maximum cross-sectional dimension is height (not width) and whose length-height regression line would not significantly differ from that of the Mohawk River population. Because of differences in shell shape, our data are also not directly applicable for Dreis- senu hugensis (quagga mussels), the other exotic dreissenid present in North America. Our length-height regression line, how- ever, is very similar to that described by Prejs et al. (1990) for zebra mussels in Poland (Fig. 4. insert). The near overlap of these regression lines suggests that our sieving data would likely be applicable to at least some European D. polymorpha populations. Sieving yielded more than three times as many 6- to 11-mm- long mussels/min as did hand picking (statistically significant, p = 0.01) (Table I). Mean yields from sieving and picking were 10.2 and 3.0 mussels/min, respectively, with 0.4% mortality after w ^ ■^ ^ Slot/! m. 1 SlotB ■^ 5,95 mm 1 N S ► \ Figure 3. Gauge used to separate zebra mussels into the following three size classes: <6, 6-11, and >11 mm in length. See text for explanation. Zebra Mussel Size Selection 749 16 14 12 E E, 'a) ■q) (/) tf) 3 '1 0 H = 0.65L' n = 1530 R^ = 0.98 10 15 20 Length (mm) Figure 4 October (1990). 16 18 20 22 24 Mussel length (mm) . Power equation expressing the relationship between shell length (L) and height (H) of W. polymorpha from the Mohawk River in 1994. Insert: Mohawk River data expressed as a simple linear equation (NYS) in order to compare it with data (Poland) in Prejs et al. 7 days for both procedures. Sieving also produced greater yields per minute than picking in the other two size classes, i.e., <6 and > 1 1 mm (statistically significant, p = 0. 10 and p = O.O.'i respec- tively) (Table 1). Sieving, however, did not completely separate the mussels into nonoverlapping size classes, as desired. Individual length mea- surements revealed that 1.1, 7.1. and 11.1% of mussels that were separated into the "ll mm" size classes were not within these ranges (Table 1). It became evident that an over- lap in the length of the mussels collected by a series of sieves was unavoidable primarily because of natural morphological variation m the zebra mussel population. Not all mussels of the same length were of the same height (Fig. 4), and height was the character a sieve pore selected for as a mussel passed lengthwise through it. Thus, two mussels of the same length, but of different heights, TABLE 1. Comparison of Hand Picking and Sieving Zebra Mussels. Mussel Length Hand (mm) Mean (±SD) Picking Sieving <6 No. of mussels collected/min 2.1 ± 0.1 6.3 ± 1.9 % Error in selection 0 1.1 ± 1.7 '7, 7-Day mortality 1.2 ± 0.4 3.2 ± 1.7 6-11 No. of mussels collected/min 3.0 ± 0.3 10.2 ± 1.4 % Error in selection 0 7.1 ± 0.3 % 7-day mortality 0.4 ± 0.1 0.4 ± 0.5 >I1 No. of mussels collected/min 2.8 ± 0.8 10.0 ± 4.4 % Error in selection 0 11.1 ± 0.2 % 7-Day mortality 0.4 ± 0.3 0.2 ± 0.4 could be retained on different sieves. Furthermore, as was ob- served in our data, the greater the variation in height for a given length, the more likely mussels of this length would be retained on different sieves. In the Mohawk River population, variation in height increased steadily with mussel length (Fig. 4), and the percent error in size class selection (1.1, 7.1, and 11.1%) also increased steadily with size class length (<6, 6-11. and >1 1 mm). In judging the usefulness of the sieving procedure versus picking, the mability of this procedure to strictly separate zebra mussel size classes by length has to be weighed in individual research projects against the significant increase in mussel yield. Perfomiing this sieving procedure in the field (Fig. 5) was actually simpler than in the laboratory. Working directly in the river eUminated the need to transport mussel-laden rocks and other substrates to the laboratory and reduced the effort required to properly dispose of contaminated materials. When a wide range of sieve pairs was used, mussels were separated into discrete ranges in length (Fig. 6). The upper and lower limits of these size ranges were determined by the pore sizes of the upper and lower sieves, respectively. As expected from natural variation in shell length and height (as discussed above), these size ranges partially overlapped each other and increased with mussel length. The data presented in Figure 6 clearly dem- onstrate that the larger the difference in pore size between the two sieves being used, the larger the range in the size of the mussels collected between them. One of the challenges in using this sieving procedure is deter- mining the exact sieve sizes that will retain mussels only within a particular length range. Individuals wanting to use sieving for zebra mussel size separation can use the regression equation, pore size = 0.47 mussel length + 0.29, as a guide for estimating the sieve pore sizes needed to collect mussels in a desired size range. 750 BlSS ET AL. E 9.50-12.50 (A 6.30-9.50 o 5.80-6.30 4.75-5.60 t 4.00-4.75 3.35-4.00 - s 2.80-3.35 0) N '« 1 2.36-2.80 2.00-2.36 1.00-2.00 •^ .-« Figure 5. Sieving under field conditions. The use of 5.6-, 2.8-, and l.O-mm sieves (top to bottom) in the Mohawk River (Crescent, NV) to collect zebra mussels, respectively, -11, 6-11, and -6 mm long. > 4 6 8 10 12 14 16 18 20 22 24 26 Mussel length (mm) Figure 6. Range in length of D. polymorpha selected by various pairs of sieves. This equation was produced from plotting and analyzing all siev- ing data produced in this study and assumes a length-height relation- ship identical to our study population (Fig. 4) and a population whose maximum cross-sectional dimension is height This equation pre- dicts, for example, that mussel size classes with approximate lengths of 1-5, 5-10. 10-15 and 15-20 mm would result from the use of pairs of sieves with pore sizes, resf)ectively, of 0.8 and 2.6 mm. 2.6 and 5.0 mm, 5.0 and 7.3 mm. and 7.3 and 9.7 mm. ACKNOWLEDGMENTS We thank William Frawley and Randy Hough (Hudson Envi- ronmental Services, Inc.) for the loan of sieves and also Barbara Griffin and Paul Molloy for their technical assistance. Critical reviews by David Oarton (Indiana University) and Victor Kennedy (University of Maryland) are gratefully acknowledged. This proj- ect was supported by funding from Project 91-18 of the Empire State Electric Energy Research Corporation. Contribution number 765 of the New York State Museum and Science Service. LITERATURE CITED Dorgelo. J. \99i. Growth and population structure of zebra mussel (Dre- issena polymorpha) in Dutch lakes differing in trophic state, pp. 79-93 In: T. F. Nalepa and D. W. Schloesser (eds. ). Zebra Mussel: Biology. Impacts and Control. Lewis Pubhshers. Boca Raton. Florida. Effler, S. W. & C. Siegfried. 1994. Zebra mussel {Dreissena polymorpha) populations in the Seneca River, New York: Impact on oxygen re- sources. Environ. Sci. Technol. 28:2216-2221. Molloy. D P. & B. Gnlfin. 1992. Biological control of zebra mussels: Screening for lethal microorganisms. J. Shellfish Res. 11:234. Prejs. A.. K. Lewandowski & A. Stanczykowska-Piotrowska. 1990. Size- selection predation by (Riitilus rutiliis) on zebra mussel (Dreissena polymorpha). Field studies. Oecologia ifieidelb.) 83:378-384. Journal of SItellfish Research. Vol. 15, No. 3, 751-758. 1996. TRENDS IN BLUE CRAB (CALLINECTES SAPIDUS RATHBUN) CATCHES NEAR CALVERT CLIFFS, MARYLAND, FROM 1968 TO 1995 AND THEIR RELATIONSHIP TO THE MARYLAND COMMERCIAL FISHERY GEORGE R. ABBE' AND CLUNEY STAGG^ ^Estiuirine Research Center Academy of Natural Sciences 10545 Mackall Rd. St. Leonard. Maryland 20685 ^Fisheries Service Maryland Department of Natural Resources 580 Taylor Avenue Annapolis, Man land 21401 ABSTRACT In an effort to understand some of the consequences of increased fishing pressure on the Maryland blue crab population, we have analyzed data collected along 12 km of western Chesapeake Bay in Calvert County from 1968 to 1995. Commercial peeler crab pots of 25-mni-pore-size mesh, baited daily with menhaden, were used to sample crab stocks at three locations with up to 60 pots fished dunng alternate weeks from June through November. Station catches were sorted, measured, and weighed by sex. From 1968 through 1995. 1 13.002 crabs were caught in 18.106 pots, of which 73% were legal size ( 127-mm carapace width). Although annual mean catch per unit effort (CPUE) vaned considerably, it appeared to be within some normal range. Total CPUE ranged from 0.85 in 1968 to 20.01 in 1981 (legal CPUE ranged from 0.73 to 13.57); Maryland commercial landings ranged from 4.7 x 10" kg to 27.1 X 10" kg during the same years. From 1968 to 1980. legal CPUE averaged 3.60; from 1981 to 1985. it averaged 8. 14; and from 1986 to 1995. it was 3.66. Thus, the legal CPUE of the most recent period was nearly identical with thai of the earliest period. There are, however, several trends that have become apparent, indicating that increased fishing pressure may be having a deleterious effect on the blue crab population. A significant correlation between this fishery-independent data and Maryland Department of Natural Resources' fisherv-dependent data demonstrates the relevance of these trends. From 1968 to 1982. the annual male percentage decreased significantly from 66 to 38% (r' = 0.79; p < 0.01 ). Since 1983. this percentage has shown greater Huctuation among years, but has shown no further decrease. The mean carapace width and weight of females have not changed significantly over time, but the mean width of males (r^ = 0.47) and the mean weight of males (r" = 0.34) have both decreased significantly (p < 0.01). Percent legal size crabs', which constituted 64-86% of the annual catch between 1 968 and 1 99 1 . has had its three lowest years ( averaging 52%) since just 1992; the percentage of legal males in the catch decreased from 56% in 1968 lo 19% in 1995 (13% in 1994) (r^ = 0.75; p < 0.01). These downward trends related to the size of males indicate that they are being removed from the population shortly after reaching legal size. With fewer large males available to crabbers, even more pressure may be exerted on females, which could eventually result in further decreases in population size and stability. KEY WORDS: blue crab. Callinectes sapidus. Chesapeake Bay. population trend, harvesting, fishery, fishing pressure, regulations INTRODUCTION men. Regulations on certain finfishes such as American shad and striped bass, which once supported heavy commercial activity, The annual value of the blue crab [Callinectes sapidus Rath- now include closed or tightly regulated seasons, changes in min- bun) landings in the Maryland Chesapeake Bay was less than that imum or maximum size, and daily and/or seasonal catch limits, of the oyster [Crossoslrea virginica Gmelin) from the late 1800s The soft clam fishery has also suffered major stock reductions in until 1983, when 23.8 x 10*" kg of crabs valued at $22.6 million recent years, resulting in increased regulations, surpassed the 3.4 x 10'' kg of oyster meats valued at $14.0 million With major reductions in the size of finfish. oyster, and clam (NMFS 1985). Since then, with the continuing decline of the fisheries, more pressure has been exerted by watermen on the blue oyster fishery due primarily to disease (MSX and Dermo). the crab to make up some revenues lost from other fisheries. Although difference between the values of the two fisheries has continued to blue crab landings have often fluctuated widely among years widen. The 1993 oyster season yielded 440 x 10' kg of oyster (Pearson 1948. Van Engel 1958, Tagatz 1965. Abbe 1983, meats worth about $3.2 million compared with 28.0 x 10* kg of Lipcius and Van Engel 1990), catches often recovered rapidly crabs (hard and soft) worth $37.4 million (NMFS 1994). The real after poor years, and the fishery has generally held up well con- difference between the size of these two fisheries is actually much sidering long-term increases in fishing pressure. Because a re- greater than indicated by these figures because the recreational source has been a consistent producer, however, is no guarantee harvest of blue crabs is also substantial and far greater than the that it will remain so forever. The striped bass (Morone saxalilus) recreational harvest of oysters. Although the recreational crab har- fishery was once thought to be inexhaustible, but poor recruitment vest adds nothing to the value of the commercial fishery, it does from the early 1970s to the early 1980s, coupled with questionable add value to the Maryland economy. The most recent recreational management practices, necessitated a moratorium on harvesting harvest expenditures (1990) were estimated in excess of $110 mil- for several years and changes in the management of the fishery lion (USFWS 1993, Stagg et al. 1994). before it began to return to a healthy condition. The oyster fishery, however, has not been the only commercial Because so many of Maryland's fisheries have declined over fishery to decline in recent years and thus affect the bay's water- the past two decades, it seems obvious why concern has been 751 752 Abbe and Stagg shown for the blue crab fishery, which is the healthiest fishery in Maryland at this time. Legislators who saw decreased crab land- ings in 1992 (14.2 x 10^ kg worth $17.6 million: NMFS 1994) were concerned that the crab fishery might also be declining with oysters, clams, and some finfish. As a result, efforts began on a series of conservation-onented changes in the fishery that were implemented before the 1994 season. These included limited entry to the fishery with a 2-y wait for a new license, limiting a com- mercial license holder to 300 crab pots with a maximum of 900 per vessel (if two additional allocations were purchased for additional crew), limiting the hours that pots could be fished, and requiring a 59-mm (2 yi<,-inch) cull ring in the upper chamber of each pot. New regulations were also set in place for the use of trotlines and crab scrapes. Further emergency restrictions were added during the 1995 season that shortened the workday by several hours, the workweek by a day. and the season by 6 wk. In the late 1960s, when many of Maryland's fisheries were far healthier than they are now, there was also concern about the possible effects of the thermal discharge from the Calvert Cliffs Nuclear Power Plant (CCNPP) (then under construction) on the blue crab population in the Chesapeake Bay waters adjacent to the plant. This led to a series of studies of population size and struc- ture from 1968 to 1983 (Abbe 1987). Those studies detected no adverse effect of the power plant on numerous parameters of the local crab populations, but after 28 y, there appear to be changes in the population structure that may be undesirable for the fishery. Although there still appear to be no station differences indicative of power plant effect, the observations of change are based on long-term trends. Long-term fishery-independent data sets of at least 10-15 y are important in understanding the dynamics of commercially valuable populations (Lipcius and Van Engel 1990). but such data .sets are also uncommon. Changes in a population over 3-5 y may be meaningful, but because of relatively short duration, it is difficult to show significance unless the changes are large. Long-term data sets, however, enable an investigator to attach significance to more subtle changes. But what do these changes at Calvert Cliffs mean? Do they represent changes in the localized population only or are they reflective of change on a broader baywide scale .' To examine this question, it was necessary to compare the long-term fishery- independent data from Calvert Cliffs with the fishery-dependent data of the Maryland Department of Natural Resources (MDNR). We felt that if the Calvert Cliffs and MDNR data compared fa- vorably over many years, then the Calvert Cliffs data might be reflective of the baywide crab population (or at least a large por- tion of it), and the changes observed at Calvert Cliffs could also be occurring elsewhere in the bay. This article reports some of the changes observed over time that may have resulted from increased fishing pressure as well as descriptive statistics of these popula- tions over the entire 28-y period. Calvert Cliffs and MDNR data are compared where possible. MATERIALS AND METHODS Calvert Cliffs Studies Stations Sampling sites were located adjacent to and on either side of the CCNPP (Fig. 1). Although the Plant Site station was located within I i) m of the plant discharge in 2.5 m of water, it did not 76° 25' W Figure L Locations of crab pots in the study area near Calvert ClifTs in central Chesapeake Bay from 1968 to 1995. receive the full effect of the thermal plume because it was off to one side. The temperature averaged l-2°C above ambient. An- other station was located in 4 m near Kenwood Beach, 8 km northwest of the plant. The third station was southeast of Rocky Point, approximately 4 km from the plant in 4 m of water. Both Kenwood Beach and Rocky Point were outside the predicted area of thermal loading from the CCNPP when they were established in 1968. Plant operation did, however, result in occasional temper- ature increases of up to 1°C at Rocky Point: Kenwood Beach was unaffected. Crab pots remained on station from June to November; they were not moved to deeper water in the fall as shallow waters cooled, as is done by many commercial crabbers. Sampling Procedures Commercial crab pots of 25-mm (I -in) galvanized-wire mesh were used to sample crab stocks at the three stations from spring until late fall, when water temperatures decreased to levels at which crabs would no longer actively enter pots (10-12°C). Be- cause blue crabs enter baited pots as a feeding response, temper- atures must be warm enough for them to actively feed. Van Engel (1962) reviewed the development of crab pots and the methods used to fish them. Most commercial pots are constructed of 38-mm ( 1 Vi-'in) mesh and will retain few crabs smaller than 76-mm (3-in) carapace width. With the present requirement for a cull ring, even much larger crabs are able to escape from commercial pots. The cull ring can be closed when catching peeler crabs for the soft crab market because 76-mni peeler crabs are legal in Maryland, in contrast to hard crabs, which must be 127 mm (5 in) across the lateral spines. The smaller meshed peeler pots used in this study (which contained no cull rings) allowed some crabs smaller than 51 mm (2 in) to be caught, although such pots were probably not very efficient at catching crabs this small. Crab pots provided an excellent means to sample the blue crab population, although they may not be the choice for sampling other crab species that exhibit agonistic behavior. Miller (1980) and Williams and Hill ( 1982) suggested that such behavior exhib- Trends in Maryland Blue Crab Populations 753 itcd by Cancer productus. Cancer irroratus. Hyas anmeus. and Scylla serrata could limit the catch in a trap because the first crab caught might then prevent others from entering. C. sapidus does not behave this way; our record for crabs in a pot set for 1 day was 37. Pots were generally fished every other week from spring until fall. In 1968, three pots were fished at each station for 5 days; from 1969 to 1981. five pots were fished for 4 days; and from 1982 to 1995, 10 pots were fished for 2 days. From 1970 to 1981, sampling began in early May and sometimes continued into De- cember. From 1982 to 1995, sampling was conducted only from early June through November. To help normalize the data, only June to November data are examined here. The crab catch is generally light during both May and December because of cold water temperatures; the elimination of May and December data reduced the total catch by 2,965 crabs or about 5.5<7( of the 54,006 crabs caught from 1970 to 1981. Pots were baited daily with menhaden, with up to 20 pots fished per station per week (15 per week in 1968), except when losses occurred to stonns or boats. Station catches were weighed by sex to the nearest 0. 1 kg, and the carapace width (CW) of each crab was measured to the nearest 'A in (3 mm) across the lateral spines. Field CW measurements were later converted to metric. The number of pots fished annually varied considerably, espe- cially during the early years, because of changes in starting dates and differences in ending dates resulting from weather and popu- lation fluctuations. Limiting the analyses to the June-November data helped reduce, but not eliminate, the variance in pots sam- pled. Bottom temperature and salinity were determined monthly by thermistor probe and titration, respectively, from 1968 to 1978 and daily during the weeks fished from 1979 to 1995, with a Beckman RS5-3 portable salinometer. Dissolved oxygen concentrations were determined monthly through 1974 and daily thereafter, either by Winkler titration or with a YSI Model 57 dissolved-oxygen meter. Data Analysis Much of the Calvert Cliffs data were examined graphically for trends and by standard linear regression techniques (Draper and Smith 1966, Kleinbaum and Kupper 1978). If a regression r^ was significant, then we concluded that the trend most likely had bi- ologic significance. MDNR Data Before 1 98 1 . data were collected from a census of commercial crabbers. On the basis of an intensive census in 1979, a random stratified sampling procedure was developed to more accurately estimate blue crab catch and effort for the pot fishery and the trotline fishery (Summers et al. 1983a, Summers et al. 1983b). For trotlines, stratification was by month; for pots, it was by month and county of residence. In 1994, MDNR returned to a census approach. Both the sampling and the census approaches have al- lowed estimates of effort and catch per unit effort (CPUE) in the pot component of the Maryland blue crab fishery since 1981. Virginia data through 1992 were compiled by the NMFS from voluntary industry reporting programs. Since 1993, the Virginia Marine Fisheries Commission has adopted a mandatory reporting system. Chesapeake Bay values are simply Maryland and Virginia values combined. Statistical Analysis Pearson correlation analysis (SAS 1985) was conducted to ex- amine the relationships among the Calvert Cliffs legal CPUE and several fishery-dependent variables. These included the reported Maryland blue crab pot catch and total catch from 1968 to 1995, the reported Virginia pot catch and total catch from 1968 to 1993, the reported Chesapeake Bay pot catch and total catch from 1968 to 1993, and the reported Maryland pot CPUE from 1981 to 1995. Because there appears to be a notable difference in the patterns of blue crab abundance between the period before 1981 and the pe- riod from 1981 onward, and because of the change in Maryland's reporting system in 198 1 , the data were thus divided and examined separately. RESULTS AND DISCUSSION MDNR The fishery-dependent data for Maryland and Virginia used in the correlation analysis are presented in Table 1 . Over the entire time period (1968-1995). the highest correlations between Calvert Cliffs CPUE and various reported catch statistics were 0.718 for the Maryland pot harvest and 0.707 for the Chesapeake Bay pot harvest (both p < 0.0001; Table 2). Although lower, there was also a significant (0.421; p < 0.05) relationship with the Virginia pot fishery. Because Chesapeake Bay numbers are totals of Mary- land and Virginia, however, they are not independent confirma- tions of the representative character of the Calvert Cliffs data. In general, the correlations for the 1968-1980 period were lower and less significant than corresponding correlations for the 1981-1995 period (Table 2). This difference could be the result of the more accurate reporting of harvests (especially for Maryland) in the latter period. In the 1981-1995 period, correlations between Cal- vert Cliffs CPUE and Maryland pot harvest (0.794) and Maryland total harvest (0.817) were both highly significant (Table 2). The highest correlation was between Calvert Cliffs CPUE and Mary- land pot CPUE (0.881; p < 0.0001) and was not unexpected. The strong relationship between these data sets demonstrates the rep- resentative character of the Calvert Cliffs data. It further suggests that when reasonably accurate measures of commercial harvests and efforts can be estimated, the results are quite similar to fish- ery-independent measures of abundance. Calvert Cliffs Since 1968, a total of 113,002 crabs have been caught in 18,106 pots at Calvert Cliffs (Table 3). The total number of crabs per pot has ranged from 0.85 in 1968 to 20.01 in 1981, whereas the number of legal crabs per pot has ranged from 0.73 to 13.57 for the same years (Table 3). Maryland commercial landings have ranged from a low of 4. 68 x 10" kg in 1968 to a high of 27. 14 x 10" kg in 1981 (Table 1). From 1968 to 1980, the total number of crabs caught per pot or caught per unit effort (CPUE) and the legal CPUE averaged 4.63 and 3.60 per year, respectively; from 1981 to 1985, these averages increased to 11.40 and 8.14, respectively; and from 1986 to 1995, they averaged 5.33 and 3.66, respectively. Because these figures indicate that the mean legal CPUE of the 1986-1995 period was almost identical with that of the 1968-1980 period, one might assume that other aspects of the populations were similar as well, but this was not so. Wide fluctuations in the size of the sample population based on CPUE continues to be the normal pattern (Fig. 2) and results in part from the fact that crabs 754 Abbe and Stagg TABLE 1. Maryland (MD), Virginia (VA), and Chesapeake Bay (CB) blue crab harvests from the pot fishery and the total fishery in millions of kilograms from 1968 to 1995; also included is the Maryland pot fishery CPUE in kilograms per pot per month from 1981 to 1994. MDPot MD Total MDPot VA Pot VA Total CBPot CB Total Year Fishery Fishery CPUE Fishery Fishery Fishery Fishery 1968 2.3 4.7 14.2 20.7 16.5 25.4 1969 6.0 11.5 10.9 16.2 16.9 27.7 1970 6.5 12.0 12.9 19.7 19.4 31.7 1971 7.1 12.5 16.2 22.0 23.3 34.5 1972 6.3 11.4 16.5 22.4 22.8 33.8 1973 5.3 9.5 12.9 17.1 18.2 26.6 1974 7.1 12.0 15.1 18.9 22.2 30.9 1975 7.2 11.8 13.9 16.1 21.1 27.9 1976 5.9 9.5 9.1 12.0 15.0 21.5 1977 6.2 9.7 14.3 17.2 20.5 26.9 1978 5.9 7.9 13.5 16.6 19.4 24.5 1979 9.2 11.7 15.0 18.5 24.2 30.2 1980 10.1 12.0 12.9 17.4 23.0 29.4 1981 16.1 27.1 25.9 14.5 19.3 30.6 46.4 1982 11.7 19.8 12.9 16.6 20.3 28.3 40.1 1983 14.3 23.8 14.9 18.3 21.2 32.6 45.0 1984 13.9 22.1 16.4 17.9 22.9 31.8 45.0 1985 17.4 26.5 19.6 15.5 17.6 32.9 44.1 1986 14.1 23.5 16.6 13.6 17.4 27.7 40.9 1987 12.7 20.9 13.9 12.2 14.8 24.9 35.7 1988 12.9 19.5 14.2 13.9 16.9 26.8 36.4 1989 11.7 19.7 13.0 15.6 20.2 27.3 39.9 1990 12.1 21.2 15.9 19.8 23.6 31.9 44.8 1991 12.9 21.8 15.5 16.4 20.4 29.3 42.2 1992 8.6 14.2 10.7 9.0 10.8 17.6 25.0 1993 16.6 26.0 13.1 20.1 24.0 36.7 50.0 1994 10.3 17.7 10.3 21.8 39.5 1995 10.2 17.0 reach legal size in only their second year (Van Engel 1958). Thus, a poor year (1970) can be followed by a good year ( 1971 ). or the reverse can occur (1986 and 1987). based on reproductive and recruitment success. Large annual fluctuations are also apparent in the Maryland and Virginia populations (Pearson 1948, Van Engel 1958, Abbe 1987, Lipcius and Van Engel 1990). Blue crab pop- ulations have remained relatively strong over the long tenn, even though the current estimate of total instantaneous mortality is about 73% and fishing mortality is about 59% (NOAA unpub- lished data). Other estimates put mortalities much higher (Roths- child et al. 1992). The apparent relative stability of the fishery in spite of high mortalities may be due to high fecundity (the external TABLE 2. Correlation coefficients (r) between Calvert Cliffs legal CPUE and various fishery-dependent catches for the pot fishery and the total fishery from Maryland (MD), Virginia (VA), and all of Chesapeake Bay (CB) for 1968-1995, 1968-1980, and 1981-1995; the Calvert Cliffs legal CPUE is also correlated with Maryland CPUE for the most recent period. I CC MD MD MD VA VA CB CB Legal Pot Total CPUE Pot Total Pot Total 1968-1995 r 0.718 0.703 0.421 0.283 0.707 0.686 P< 0.0001 0.0001 0.032 0.153 0.0001 0.0001 n 28 28 26 27 26 27 1968-1980 r 0.691 0.725 0.224 -0.006 .597 0.427 P< 0.009 0.005 0.461 0.984 0.031 0.146 n 13 13 13 13 13 13 1981-1995 r 0.794 0.817 0.881 0.321 0.298 0.589 0.645 P< 0.0004 0.0002 0.0001 0.285 0.300 0.034 0.013 n 15 15 14 13 14 13 14 Trends in Maryland Blue Crab Populations 755 TABLE 3. Catches of blue crabs in Chesapeake Bay near Clavert CHfTs from 1968 to 1995. Total Total Legal Legal Legal No. No. Legal Total Crabs Crabs Males Females Total of of Percent Legal Legal .Size Total Percent Pots per per per per Year No. Males Females Males Males Females 1 127 mm) Sublegal Legal Fished Pot Pot Pot Pot 14h8 239 158 81 66.1 133 73 206 33 86.2 281 0.85 0 73 0.47 0.26 ]1(t<> 2.833 1.995 838 70 4 1 .450 556 2.006 827 70.8 472 6.00 4 25 3.07 1.18 1470 1.318 800 518 60.7 613 400 1.013 305 76.9 504 2.62 2.01 1.22 0.79 1471 4.463 2,461 2.002 .55 1 1.811 1.632 3.443 1,020 77.1 590 7.56 5.84 3,07 2.77 1472 2.699 1,611 1.088 59,7 1.160 906 2.066 633 76.5 690 3.91 2.49 1 68 1.31 1473 2.903 1.663 1.240 57.3 1.264 1,026 2.290 613 78,9 727 3.99 3 15 1 74 1.41 1474 3,718 2.225 1.493 59,8 1.658 1.137 2.795 923 75.2 640 5 81 4 37 2.59 1.78 1475 4.467 2.142 2.325 48.0 1,713 2.042 3,755 712 84.1 751 5 45 5 00 2.28 2.72 1976 2,735 1.192 1 .543 43.6 743 1.144 1.887 848 69.0 734 3 73 2 57 1.01 1.56 1977 1,998 1.024 974 51 3 883 818 1,701 297 85.1 6.30 3 17 2.70 1.40 1.30 1978 3.340 1.622 1.718 48 6 1.155 1.356 2.511 829 75.2 740 4 51 3.39 1.56 1.83 1974 5,386 2.839 2.547 52.7 2.109 2.129 4.238 1,148 78.7 699 7 71 6,06 3.02 3.05 1980 3.206 1.361 1 .845 42 5 1.089 1 ,653 2.742 464 85.5 741 4.33 3 70 1.47 2.23 1981 14.809 6.691 8.118 45.2 3.385 6,657 10.042 4,767 67.8 740 20.01 13.57 4.57 9.00 1982 3.797 1.449 2.348 38.2 981 1.920 2.901 896 76.4 657 5.78 442 1.49 2.92 1983 5.400 2,640 2.760 48.9 1.780 2.446 4.226 1,174 78.3 682 7.92 6.20 2.61 3.59 1984 7.347 3,470 3.877 47.2 1.736 3.190 4.926 2,421 67.0 653 11.25 7.54 2.66 4.89 1985 7.373 2.870 4.503 38.9 1.497 4.015 5.512 1,861 74.8 613 12.03 8,99 2.44 6.55 1986 3.823 1 .548 2.275 40.5 1.004 2.088 3.092 731 80.9 620 6.17 4.99 1.62 3.37 1987 1.573 880 693 .55.9 570 547 1.117 456 71.0 687 2.29 1.63 0.83 0.80 1988 3.380 1.620 1.760 479 483 1.435 2.418 962 71.5 655 5.16 3.69 1.50 2.19 1989 4.287 1.790 2.497 41 8 759 1.991 2,750 1,537 64.1 684 6.27 4.02 1.11 2.91 1990 3.474 1.458 2,016 42.0 1 ,080 1.824 2,904 570 83.6 662 5.25 4.39 1.63 2.76 1991 4.073 1.260 2,813 30.9 771) 2.627 3,397 676 83.4 678 6.01 5.01 1.14 3.87 1992 2.802 1.418 1 ,384 .50 6 535 1.042 1,577 1,225 56.3 618 4.53 2.55 0.87 1.69 1993 5.703 2.621 3,082 46.0 1 ,367 2,586 3,953 1.750 69.3 687 8.30 5.75 1.99 3.76 1994 3,521 1.693 1,828 48.1 466 1,170 1.636 1,885 46.5 612 5.75 2.67 0.76 1.91 1995 2,335 1.271 1.064 54.4 452 809 1,261 1,074 54.0 659 3.54 1.91 0,69 1.23 Total 113,002 53.772 59.230 33,146 49.219 82,365 30,637 18.106 Mean 4036 1920 2115 47.6 1184 1758 2942 1094 72 9 647 6 24 4.55 1.83 2.72 egg mass of a female may contain from 750,000 to 8 million eggs, depending on her size; Prager et al. 1990) and a short life span of 2-3 y (Hay 1905, Churchill 1919. Van Engel 1958). Although recent evidence based on tag returns (McConaugha 1991 ) indicates that blue crabs may live to be 7 to 8 y old, most probably do not live longer than 2 or 3 y because they are exploited by the fishery. Ultimate age may be critical in a mathematical population model, but it is less critical to the crab population itself because blue crabs are capable of spawning in their second or third year. If there was 20 18 1fi 1- o Q. 14 . rr LU 12 a. w 10 ? ; ■; / ,' 1 IT 8 - \ A U A Total ,\ / \ b " A /A / -^X / ' V y •\ / ■ * \/ ' ' \ 4 /• \ /,' ',\ y r 'A i \ ''"'' \ ■ '■■ ^ 2 /• ■: Legal size '*■-•' ■ ' » . . . . 1 .... 1 1970 1975 1980 1985 1990 1995 Figure 2. Total and legal-size CPUE near Calvert Cliffs from 1968 to 1995. considerable reproductive output from 4- to 7-y-old blue crabs in the Chesapeake Bay. then the harvest of 2 and 3 y olds could have a major effect on population size, but this is probably not the case. The 53.760 males caught between 1968 and 1995 accounted for slightly less than half of the catch (47.6%), or 49.5% if aver- aged across equally weighted years. The trend for male percents has been downward throughout much of this study, and the decline is significant across all years from 1968 to 1995 (r" = 0.476; p < 111 UJ O m 80 70 60 50 30 y=-410 1+0 229X R-"=0 017 Y=3654 8-1 824X R-=0 788 p<0 001 1970 1975 1980 1985 1990 1995 Figure 3. Annual percentage of total catch near Calvert Cliffs con- sisting of males showing the decline from the late 1960s through the early 1980s and the subsequent leveling off. 756 Abbe and Stagg 0.001). It was most evident, however, between 1968 and 1982 (Fig. 3), when males decreased from 66.1 to 38.2% of the total catch (r^ = 0.788; p < 0.001). Because the ratio of females to males is higher at higher salinities, salinity data from 1968 to 1981 were analyzed to determine if there had been an increase in salinity over time that would explain the decline in males, but no increase in salinity was apparent (Abbe 1983). From 1983 to 1995. the percentage of males fluctuated more than it had during earlier years, but there was no further decline (r^ = 0.017). When percent legal males was examined, however, the decline was contmuous over time, and this is vital to the fishery because legal crabs are the ones that comprise the fishery. The three high- est years were in 1968-1970 (as with total males), with a signif- icant decrease until 1995 (r* = 0.753; p < 0.001); however, unlike percent total males, which began to level off in the early 1980s (Fig. 3), percent legal males continued to decline, reaching a low of just 13% in 1994 (Fig. 4). In contrast to legal males, however, legal females have shown a significant increase (r" = 0.233; p < 0.01 ) over the entire period (Fig. 4). but it was far less significant than the decrease of legal males. Legal females, how- ever, have shown lower percents during the last 4 y than during many earlier years, and this may be the result of added fishing pressure as well. The 82,365 legal crabs accounted for 72.9% of the total, 73.7% when averaged across equally weighted years. From 1968 to 1991, legal crab percents formed a fairly tight group, ranging from 64 to 86% of the annual catch. During 7 of these years, at least 80% of the catch was legal size, and during 15 y, at least 75% was legal. A regression line fit to the 1968-1991 data is flat (r^ = 70 60- 50 40 05 00 O _l < LU HI o Qi LU Q. 30 20 60 50 40 30 20 10 FEMALES Y = -1058.81 +0 56X R' = 0 233 p<001 MALES Y = 2379 13- 1 18X R' = 0 753 p<0.001 0.018). However, from 1992 to 1995, legal crabs fell to 56, 69, 46, and 54% of total catch, respectively. When these years are included in the data set, the regression becomes significant (r = 0.263; p < 0.01 ). Although the regression is significant, a smooth curve illustrates the changes better than a straight line (Fig. 5). The fact that the three lowest percentages occurred during the last 4 y is another indication that fishing pressure may be greater than the legal portion of the population can withstand. During the 28 y, male crabs averaged 134-mm CW and 154 g while females averaged 146-mm CW and 149 g. Females have longer lateral spines than males and thus are lighter per unit width than males, as indicated in Figure 6 and by others (Ncwcombe et al. 1949, Tagatz 1965, Pullen and Trent 1970). These means indicate only that males were smaller and heavier than females, but they do not illustrate what has happened to the male popula- tions. Male crabs have shown a decrease in mean size (both width and weight) over time in contrast to females. The regression line for female width is almost flat (r^ = 0.021), although annual means have been below average the last 4 y (Fig. 7). Males, however, decreased significantly (r~ = 0.471; p < 0.001), and 7 of the last 8 y have been below the long-term average (Fig. 7). Females showed no decrease in weight overtime (r = 0.015), but males decreased significantly (r* = 0.344; p < 0.01). Because width and weight are highly correlated (p < 0.001 for both sexes; Fig. 6), it is not surprising that both width and weight show similar trends by sex. The decrease in male size led to analysis of just the legal sector of the population because the decline could have resulted simply from increased numbers of sublegal males, which could have driven down the mean size of the entire male population. If this were so, there might be little effect on the mean size of the legal males, which make up a major part of the fishery. The regression of the mean width of legal males did show a significant downward trend (r^ = 0.356; p < 0.001) (Fig. 8), similar to that for the entire male population (Fig. 7). The regression for legal females was almost flat (Fig. 8), as it was for the total female population (Fig. 7). The annual mean widths of females were always greater than for males, but weights were not. The annual mean weights of males were greater than those of females from 1968 to 1980 and during 17 of the 19 y from 1968 to 1986. Since 1987, however, males have been heavier than females only once (1988), further 1970 1975 1980 1985 1990 1995 I Figure 4. Percentages of legal males and females relative to the total catch near Calvert Cliffs from 1968 to 1995. 1970 1975 1980 1985 1990 1995 Figure 5. Annual mean percentage of legal crabs relative to the total catch, showing the stable period from 1968 to 1991 and the low levels during 3 of the last 4 y. Trends in Maryland Blue Crab Populations 757 • 200 Males* • / O) 180 Y = 2 73 X R" = 0 708 -211 15 / / o 1- ./ o / T • K o / O 160 • • • • • 1 0 / ^ , < • o /6 o n- 140 y/ • /» o° ° 0 o / • o / r° i?n / Females n / • . / Y = 2 47 X- 210 83 = 0 772 , 1 1 110 115 120 125 130 135 140 145 150 155 160 CARAPACE WIDTH (mm) Figure 6. Relationship between annual mean CW and weight for males and females caught near Calvert Cliffs from 1968 to 1995. supporting the decrease in average male size. Tlie cause of this decline in male size is not certain, but it may well be associated with increased fishing effort. The more crabs that arc removed from the population shortly after reaching minimum legal size, the fewer there are that can attain larger size. As fishing pressure mcreascs. the situation only gets worse, although there is a lower limit that can be reached because of the 127-mm minimum legal size limit. Females may be somewhat immune to this because they often initiate their terminal molt at a sublegal size of about 1 15- riim CW (Knotts 19X9) and average ISS-nim CW when finished (Knotts 1989, Mines et al. 1987), although many get much larger. The ultimate size of females is probably influenced more by events occurring earlier in development than by environmental conditions at the time of the terminal molt (Haefner and Shuster 1964). This jump to larger size made by maturing females is fortuitous for the overall population because fecundity is directly related to CW (Prager et al. 1990). Thus, if the mean size of females was de- creasing as is occurring with males, the number of eggs produced by females would be decreasing also. There has been some concern that the size of the fall population of females (those migrating down bay to spawn the following spring) may be decreasing because of more intense fishing pres- F Females F IRO Y = 341 - 0 098X ^-^ R' = 0,021 1 1- n O 150 ° . i.__^i_\_:^^__ LU o ^ „ „ ° ° ° „^ o o • o ■ © < 140 •^ — ^-*-L r n ^ ^: . ^ • '~~~-~—^^^ • • o o 130 ^^^^^^^ III Males* • • • • 2 12U Y= 1448-0 663X R' = 0 471 p < 0 001 . 1 .... 1 . . • . . 1 .... 1 .... 1 .... 1 1970 1975 1980 1985 1990 1995 E E, (/) CO o < o ID o X 170 160 150 140 160 130 FEMALES Y = 362 30-0 105X R' = 0 047 MALES UJ • C) • < Q. ^ 150 _ ^ • ■ • < * '"^'^---^ O • z • < • UJ ^ 140 - Y = 937.37- 0.399 X R' = 0 356 P = 0.001 Figure 7. Annual mean CW for all crabs, showing the relative stabil- ity of females and the declining size of males. 1970 1975 1980 1985 1990 1995 Figure 8. Annual mean CW of legal males and females. sure and because fishing continues later into the year than it once did because of the collapse of other fisheries (personal observa- tions). When the oyster fishery was healthy, many crabbers began oystering as early as September or October. With the collapse of the oyster fishery, however, some crabbers worked far into De- cember because little fishery activity remained for them after crab season, although this was not possible in 1995 when emergency restrictions ended the season in mid-November. If the size of the fall population of females has been relatively stable across years, then CPUE should have decreased if effort by commercial crab- bers has increased. If the fall population size of females has been decreasing, then the CPUE should have decreased at an even greater rate. Data on female CPUE during October and November do not. however, present overwhelming evidence that population sizes of fall females have been declining (Fig. 9), although 1992, 1994, and 1995 were three of the lowest years since 1988. Figure 9 represents legal females only, but the same pattern exists when all females are included because 92% of the females caught in October and November have been legal size. The facts that the numbers and sizes of male crabs have de- clined in relation to females and that the percentage of the catch composed of legal crabs has declined during a time of increased fishing pressure indicate a certain amount of stress on the popu- lation, but these data do not point to a collapse of the Maryland blue crab fishery. Some of the above data may even point to a fishery in fairly good condition. There are, however, enough warning signs of fishery instability that monitoring of certain pop- ulation parameters should be continued, and acted on if necessary, to prevent further deterioration of the fishery, as has occurred with other fisheries in the Chesapeake Bay. 758 Abbe and Stagg a. 1970 1975 1980 1985 1990 1995 Figure 9. Legal CPUE for males and females and the percentage of legal females relative to the total catch during October and November 1968-1995. Many watermen who depend on blue crabs for a livelihood are reluctant to admit that changes have occurred in the structure of the populations and therefore see no need for further regulations, al- though It may be regulations that extend their careers. Many crab- bers also feel that there is little relationship between localized scientific data and bay wide crab populations, even though samples have been collected from the same populations that they are fish- ing. The relationship exists and has been demonstrated here. A rebound in some of these population parameters during the next year or two might be viewed as a sign of recovery that could force managers to relax regulations. However, the trends exhibited here did not develop over 1 or 2 y; rather, they developed over many years and will need to be monitored for many more years before we can be assured that downward trends have leveled or been reversed. If downward trends continue, additional regula- tions may be required. ACKNOWLEDGMENTS We thank all of the individuals who assisted in the collection of data at Calvert Cliffs during the 28 years of this project, but especially R. Cantin, M. Newman. W, Yates, Jr., B. Albright, and J. Hixson. Major funding for the Calvert Cliffs study was from the Baltimore Gas and Electric Company. LITERATURE CITED Abbe, G. R. 1983, A study of blue crab populatmns in Chesapeake Bay In the vicinity of the Calvert Cliffs Nuclear Power Plant, IQhS-lQXl J Shellfish Res. .3:I83-19.V Abbe, G. R. 1987 Blue crabs pp 12.VI42 In: K. L. Heck, Jr. (ed). Ecological Studies in the Middle Reach of Chesapeake Bay; Calvert Cliffs. Springer- Verlag, New York. Churchill, E. P., Jr. 1919. Life history of the blue crab Bull U.S. Bur. Fish. 1917-1918. 36:91-128. Draper, N R. & H Smith. 1966. Applied Regression Analysis John Wiley and Sons, Inc. New York. 407 pp. Haefner, P A , Jr. & C. N. Shuster, Jr. 1964. Length increments during terminal molt of the female blue crab, Callinectes sapidus. in different salinity environments. Chesapeake Sci. 5:114—118. Hay, W, P. 1905. The life history of the blue crab [Callinectes sapidus). U.S. Bur Fish. Rep. 1904:395^13 Hines, A., R. Lipcius & A. Haddon 1987. Population dynamics and habitat partitioning by size, sex. and molt stage of blue crabs Calli- nectes sapidus in a subestury of central Chesapeake Bay. Mar. Ecol. Prog. Ser. 36:55-64. Kleinbaum, D. G. & L. L. Kupper. 1978 Applied regression analysis and other multivariable methods. Duxbury Press, North Scituate, MA. 556 pp. Knotts, K. 1989 Preliminary Stock Assessment of the Chesapeake Bay Blue Crab Population. Thesis submitted to the University of Maryland. College Park, MD. 206 pp Lipcius, R. N. & W. A. Van Engel. 1990. Blue crab population dynamics in Chesapeake Bay: variation in abundance (York River, 1972-1988) and stock-recruit functions. Bull. Mar. Sci. 46:180-194. McConaugha, J. 1991. Tag-recapture study of the spawning stock of the Chesapeake Bay blue crabs. Tech. Rept. 91-1. College of Sciences, Old Dominion University, Norfolk, VA. 29 pp. Miller, R. J. 1980. Design criteria for crab traps J. Cons. Inl. E.xp. Mer. 39:140-149. National Marine Fisheries Service (NMFS). 1985 and 1994. Preliminary commercial fisheries landings, by state (Maryland 1983 and 1993). U.S. Dept. Comm., Nad. Mar. Fish. Serv , Res Stat. Div., Wash- ington, DC. Newcombe, C. L., F. Campbell & A. M. Eckstine. 1949. A study of the form and growth of the blue crab Callinectes sapidus. Growth 13:71- 96. Pearson, J. C. 1948. Fluctuations in the abundance of the blue crab in Chesapeake Bay. U.S. Fish Wildl. Ser\-.. Res. Repl. 14. 26 pp. Prager, M. H , J. R. McConaugha, C. M. Jones & P. J. Geer. 1990. Fecundity of blue crab, Callinectes sapidus, in Chesapeake Bay: bio- logical, statistical and management considerations. Bull Mar. Sci, 46:170-179. Pullen, E. J & W. L. Trent. 1970. Carapace width-total weight relation of blue crabs from Galveston Bay, Texas. Trans. Am. Fish. Soc. 99:795- 798. Rothschild, B., J. Aull, E Patrick, S Smith, H. Li, T. Maurer, B. Daugherty, G. Davis, C. Zhang & R. McGarvey. 1992. Assessment of the Chesapeake Bay blue crab slock. Univ. Maryland, Ches. Biol. Lab. CB92-003-036, CEES 07-4-30307, Solomons, MD. SAS Institute, Inc. 1985. SAS Procedures Guide for Personal Computers, version 6. SAS Institute, Inc., Cary, NC. Stagg, C, M. Holloway, L. Rugulo, K. Knotts & L. Kline. 1994. Eval- uation of the 1990 recreational, charter boat, and commercial striped bass fishing surveys, and design of a recreational blue crab survey. Ches. Bay Res. Monitor CBRM-FR-94-1, MDNR, Annapolis, MD. Summers, J. K., H. W Hoffman & W. A. Richkus. 1983a. Randomized sample surveys to estimate annual blue crab harvests by a multi-gear fishery in the Maryland waters of Chesapeake Bay. A'. Am. J. Fish. Manag. 3:9-20. Summers, J. K., W A. Richkus, H. W. Hoffman, C. Bonzek, H. H. King & M Burch 1983b. Application of random sample survey de- signs to estimate the commercial blue crab harvest in Maryland. N. Am. J. Fish. Manag. 3:21-25. Tagatz, M. E. 1965. The fishery for blue crabs in the St. Johns River, Florida, with special reference to fluctuation in yield between 1961 and 1962. U.S. Fish Wildl. Ser\-. Spec. Sci. Rep. Fish. 501. 11 pp. United States Fish and Wildlife Service (USFWS). 1993. The 1993 Na- tional Survey of Fishing, Hunting and Wildlife-Associated Recreation. U.S. Dept. of the Interior, Washington, DC. Van Engel, W. A. 1958. The blue crab and its fishery in Chesapeake Bay. I. Reproduction, early development, growth, and migration. Comm. Fish. Rev. 20:6-17. Van Engel, W. A. 1962. The blue crab and its fishery in Chesapeake Bay. II. Types of gear for hard crab fishing. Comm. Fish. Rev. 24:1-10. Williams, M. J. & B. J. Hill. 1982 Factors influencing pot catches and population estimates of the portunid crab Scylla serrata. Mar. Biol. 71:187-192. 1 J ounuit ,>f Shellfish Research . Vol. 15. No, 3. 75'^-761. 1996. EFFECT OF NITRITE ON GROWTH OF JUVENILE RED SWAMP CRAWFISH, PROCAMBARVS CLARKII HUI LIU AND JAMES W. AVAULT, JR. School of Forestiy. Wildlife and Fisheries Louisiana Agricultural Experiment Station Louisiana State University Agricultural Center Baton Rouge. Louisiana 70803 Crawfish is a traditional culture species in Louisiana, dating back to the 1870s (de la Bretonne 1970). Crawfish aquaculture is one of the most important cultured crustaceans in the United States, with annual production near 60 million pounds (USDA 1993). The major culture species are red swamp crawfish. Pru- cambarus clarkii. and white river crawfish, Procambarus zonan- guhis. Crawfish production is affected by water quality, forage type and abundance, and stocking density (Lutz and Wolters 1986, de la Bretonne and Romaire 1987, de la Bretonne and Romaire 1989, Avault and Brunson 1990). Nitrite is a limiting factor in aquaculture production (Colt and Armstrong 1981 ). High concen- trations of nitrite are lethal to culture species, and a sublethal concentration may affect growth (Colt and Armstrong 1981 , Chien 1992). Hymel (1985) and Gutzmer and Tomasso (1985) investi- gated the acute toxicity (96-h/LC5„) of nitrite to red swamp craw- fish. Our preliminary study evaluated the chronic effect of nitrite on juvenile red swamp crawfish. Juvenile red swamp crawfish were obtained from a drainage canal at the Aquaculture Research Laboratory, LSU Agricultural Center, Baton Rouge, LA. Crawfish were held in a recirculating system that contained dechlorinated Baton Rouge city water (also used as dilution water) for at least 1 wk before the test. Water hardness was adjusted to approximately 100 mg/L as CaCO, with calcium chloride. The chloride concentration of dilution water was 78.9 mg/L, which might slightly increase the tolerance of the crawfish to nitrite. The laboratory temperature was maintained at 22.0 ± 2.0°C, and the photoperiod was 16-h light and 8-h dark with fluorescent light. A 1 ,000 mg/L stock solution of nitrite was made by dissolving 4.93 g of reagent grade sodium nitrite in a 1-L volumetric flask and then bringing it up to volume with distilled water. Test solu- tions (0.59, 2.97, and 4.75 mg/L) were prepared by adding an appropriate amount of stock solution to a 24-L tank and then bringing it up to volume with dilution water. Water quality (tem- perature, dissolved oxygen, pH, and ammonia) was monitored before the test solution was distributed randomly to triplicate tanks and during the test. Test containers were 40.5-L plastic tanks. Crawfish were measured, weighed, and placed separately into 9-cm polyvinyl chloride pipe chambers. One end was covered with plastic mesh netting to prevent cannibalism. There were 8 crawfish per replicate, 24 crawfish per treatment. A control, which did not contain any test substance, was maintained concurrently with the three test solutions. Tanks were continuously aerated, and test solutions were renewed daily to maintain proper concentrations. Crawfish were fed with a commercial trout diet (40% protein) once daily at 1-2% of body weight. Dead crawfish, exuviae, and un- eaten feed were removed daily when the test solution was re- newed. Nitrite concentrations were monitored twice per week with an ORION 960 ion-electrode analyzer (HNU Systems, Inc.). The test lasted 30 days. Wet weight and total length were recorded at the initiation of the test and thereafter every 15 days. Mortality and observed sub- lethal effects were also recorded. An analysis of variance was used to evaluate the chronic effects of nitrite among different concen- trations. Growth differences in concentration means were declared significant at p < 0.05 with Duncan's new multiple range test. A Statistical Analysis Software (SAS) program was used for data computation and analysis (SAS 1992). The growth (mean wet weight and total length) of the crawfish exposed to each test solution is shown in Table 1 . Mean crawfish growth rates were L81, 1.55, 1.08, and 1.23 g in weight or 8.7, 7.1, 4.0, and 5.8 mm in length at concentrations of 0, 0.59, 2.97, and 4.75 mg/L of nitrite-N, respectively, over the 30-day period. The percent weight gain and percent length increase of juvenile crawfish exposed to each test solution are presented in Table 2. No statistical difference in weight and length gain was observed on day 15, but after 30 days, the weight and length gains of crawfish exposed to 2.97 and 4.75 mg/L of nitrite-N were significantly lower than controls and the 0.59 mg/L treatment (p < 0.05). Nitrite reduced the growth of the crawfish. Crawfish mortality among different concentrations was not significantly different (p > 0.05). However, more crawfish were observed dead in test solutions during molting than from cannibalism (Table 3). This might have been caused by nitrite toxicity. Nitrite concentrations were maintained within 75-125% of the nominal test concentra- tions (Table 4). Dissolved oxygen, pH, temperature, and ammonia were monitored within suitable ranges for crawfish growing during the experiment. Nitrite is known to be toxic to freshwater crawfish; however, literature about the acute and chronic effects of nitrite to crawfish is limited. Gutzmer and Tomasso ( 1985) found that the 96-h LC50 of nitrite to adult red swamp crawfish was 28 mg/L of nitrite, and Hymel (1985) reported that the acute toxicity of nitrite (LC50) to juveniles was 5.9 mg/L of nitrite-N. Hymel (1985) also suggested that nitrite concentrations less than 1.0 mg/L of nitrite-N should have no negative effect on the production of crawfish. Beitinger and Huey ( 1981 ) reported the 96-h LC.^,, for Procambarus simu- lans to be 1 .9 mg/L of nitrite-N, and Johnson (1983) reported the 48-h LC50 for Procambarus acutus to be 600 mg/L. Our study showed that high concentrations of nitrite O0.59 mg/L) may sig- nificantly affect the growth of juvenile red swamp crawfish. Com- pared with penaeid shrimp and freshwater prawns, red swamp crawfish are relatively sensitive to nitrite (Armstrong et al. 1976, Wickins 1976, Jayasankar and Muthu 1983, Chen and Chin 1988, 759 760 Liu and Avault TABLE 1. Mean weight and length (standard deviation) of juvenile red swamp crawflsh, P. clarkii, exposed to different concentrations of nitrite for 30 days. Nitrite-N (mg/L) Odav Weight (g) 15 dav 30 dav 0 day Length I mm) 15 day 30 day 0.00 0.59 2.97 4.75 1.45 2.11 (0.46) (0.64) 1.28 1.97 (0.44) (0.69) 1.49 2.04 (0.37) (0.57) 1.38 1.92 (0.49) (0.87) 3.26 40.1 (0.90) (4.5) 2.84 39.3 (1.34) (4.7) 2.57 41.4 (0.88) (4.2) 2.61 40.3 (0.99) (5.5) 43.4 (4.6) 41.8 (4.7) 42.8 (3.5) 41.9 (6.0) 48.8 (4.6) 46.5 (6.7) 45.4 (5.6) 46.0 (6.0) TABLE 2. Mean weight and length gain of juvenile red swamp crawfish, P. clarkii, exposed to different concentrations of nitrite for 30 days" Nitrite-N (mgA.) Weight Gain ( % ) 15 day 30 day Length Gain ( % ) 15 dav 30 day 0.00 0.59 2.97 4.75 46.7" 51.8" 37.2" 37.1" 124.4" 121.4" 72.5'= 89.9' 8.1" 6.0" 3.6" 3.8" 21.7" 18.5"' 10.0' 14.8"-' ' Data in the same column having different superscripts are significantly different (p < 0.05). Note: length/weight gain 1%) = (mean length/weight at time t - mean initial length/wt) x 100/(mean initial length/wt) TABLE 3. Mortality of juvenile red swamp crawfish, P. clarkii, exposed to different concentrations of nitrite for 30 davs. Nitrite-N Mortality ( % ) (mg/L) Cannibalism Molting Total 0.00 25.0 0.0 25.0 0.59 12.5 16.7 29.2 2.97 4.2 20.8 25.0 4.75 8.3 25.0 33.3 Chen et al. 1990a. Chen et al. 1990b. Chen and Lei 1990, Chen and Tu 1990). Hymel (1985) repoiled that high nitrite concentra- tions were unhkely in crawfish ponds because of the extensive nature of the cultivation practice in which no formulated feeds are used. However, further study should be conducted to identify the long-term effects of nitrite on the growth of crawfish. High nitrite concentrations may be a problem m soft-shell crawfish operations that use recirculating systems. Further research is needed to eval- uate the toxicity of nitrite at different levels of chloride. I TABLE 4. Measured nitrite-N concentrations in test containers with juvenile red swamp crawfish, P. clarkii, during 30-day exposure. Nominal Nitrite (mg/L) Measured Concentrations (i mg/L) Week 1 Week 2 Week 3 Week 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.59 0.59 0.59 0.59 0.58 0.69 0.72 0.52 0.66 2.97 2.63 2.80 2.67 2.68 2.51 2.78 2.96 2.84 4.75 4.20 4.36 4,45 4.45 4.09 4.38 4.71 466 Effect of NrrRiTE on P. cl\rkii 761 LITERATURE CITED Armstronj;. D. A.. M, J, Stephenson & A. W. Knight, 1476. Acute tox- icity of nitrite to iai-vae of the giant Malaysian prawn. Miunihruchinm rosenheifiii. Aqiiaciilnire 9:39-46. Avaull. J. W. & M. W. Brunson. 1990. Crawfish forage and feeding systems. Rev. Aqiial. Sci. 3; 1-10. Beitinger. T. L. & D. W, Huey. 1981 . Acute to.xicity of nitnte to crayfish Prociimharus simiiUiits in vaned environmental conditions. Environ. Polliil. 26A:305i-3 1 1 . Chen, J. C. & T. S. Chin. 1988. Acute toxicity of nitrite to tiger prawn, Penaeus monodon larvae. Aquaculture 69:253-262. Chen, J. C. & S. C. Lei. 1990. Toxicity of ammonia and nitnte to P<'- nafu.t monodon juveniles. 7. World Aquacult. Soc. 21:300-306. Chen, J. C, P. C. Liu & S. C. Lei, 1990b. Toxicity of ammonia and nitrite to Penaeus monodon adolescents. Aqiiaciihiire 89:127-137. Chen, J. C, Y. Y. Ting. J. N. Lm & M. N. Lin. 1990a. Lethal effects of ammonia and nitrite on Penaeus cbinensis juveniles. Mar. Biol. 107: 427^31. Chen, J. C. & C. C. Tu. 1990. Acute toxicity of nitrite to larval Penaeus japonicus. J . Fish. Soc. Taiwan 17:277-287. Chien, Y. 1992. Water quality requirements and management for marine shrimp culture, pp. 144-156. In: James Wyban (ed.). Proceedings of the Special Session on Shrimp Fanning, World Aquaculture Society, Baton Rouge, LA. Colt, J. E. & D. A. Armstrong. 1981 . Nitrogen toxicity to crustacean, fish and molluscs, pp. 34-47. In: C. J. Allen and E. C. Kinney (eds.). Proceedings of Bioenginecring Symposium for Fish Culture, Fish Cul- ture Section, Amencan Fishenes Society, Bethesda, MD. de la Bretonne, L. W. 1970. History of the crawfish industry. Unpub- lished. de la Bretonne, L. W & R. P Roniaire 1987. Crawfish Culture— Site Selection, Pond Construction and Water Quality. Louisiana Coopera- tive Extension Service 87-CRSR-2-3218. Louisiana State University Agricultural Center, Baton Rouge, LA. de la Brotonne, L. W & R P. Roniaire, 1989. Commercial crawfish cultivation practices: a review. ./. Shellfish Res. 8:267-275, Gutzmer, M. P. & J. R. Tomasso, 1985, Nitnte toxicity to the crayfish Procamharus clarkii. Bull. Environ. Contam. Toxicol. 34:369-376. Hymel, T. M. 1985. Water Quality Dynamics in Commercial Crawfish Ponds and Toxicity of Selected Water Quality Variables to Procam- harus clarkii. Master's thesis. Louisiana State University, Baton Rouge. LA Jayasankar, P. & M. S. Muthu. 1983. Toxicity of nitnte to the larvae of Penaeus mdicus. Ind J. Eish. 30:231-240. Johnson, S. K. 1983, Water Quality in Crayfish Farming. Texas Agncul- tural Extension Service, Fish Disease Diagnostic Laboratory Circular FDDL-513 Lutz, C. G. & W. R. Wolters. 1986. The effect of five stocking densities on growth and yield of red swamp crawfish Procamharus clarkii. J. World Aquacult. Soc. 7:33-36. SAS. 1992. SAS User's Guide. SAS Institute, Gary, NC. USDA. 1993. Aquaculture Situation and Outlook. Commodity Economics Division, Economics Research Service, USDA, WIckins, J, F. 1976. The tolerance of warm-water prawns to recirculated water. Aquaculture 9:19-37. I Journal of Shellfish Research. Vol 15. No, 3. 763-768. 1996. NOTES ON THE BLUE CRAB FISHERY IN THE APALACHICOLA, FLORIDA ESTUARY DONNA G. INGLE AND ROBERT M. INGLE Adehmto Corporation Tallahassee . Florida ABSTRACT Harvested, cooked blue erahs were inlermiltently sampled and measured in a processing house over a period of 2 y. Information was gathered in the range of sizes, means, and modes of males, inmiature and mature females, and gravid females. KEY WORDS: Blue crah, Callinecles sapiJiis. female maturity, population variation, sizes of sexes, estuarine dynamics INTRODUCTION Callinectes sapidiis Rathbun 1896. a swimming crab, belongs to the family Porliinidaf. Its Itfc history has been mvcstigated by Barnes (1904). Hay (1905), Churchill(1921), Truitt (1939), Dar- nell ( 1959), and Tagatz (1968). An excellent summary is provided by the Gulf States Marine Fisheries Commission (1970). Mating occurs in brackish water areas, where males remain throughout their life cycle. After fertilization, females migrate to more saline waters where spawning occurs (Tagatz 1968). Most spawning takes place in spring and summer. Larvae migrate to brackish water estuaries to mature. Growth occurs by eedysis. This process is the basis of the soft crab fishery, an industry and its potential in Florida described by Otwell et al. ( 1980). Many studies have shown inshore-offshore migrations of blue crabs from various areas: Van Engel (1958), Darnell (1959), Wa- terman ( 1961 ), Fischler and Walburg (1962), Tagatz ( 1968), Judy and Dudley (1970), Gulf States Marine Fisheries Commission (1970). Oesterling (1976). however, found indications of along- shore migrations on the western coast of Florida in the Gulf of Mexico. His tagging studies indicated that all nonlocal movement on the western coast was in a northerly direction along the pen- insular portion of the state and westerly along the panhandle. Oesterling and Adams (1982) theorized that the Apalachicola es- tuarine system is the primary spawning ground for blue crabs on Florida's Gulf Coast. Oesterling's study was conducted for only a short time (less than a year). More tagging by others followed. The later tagging confirmed earlier findings of a long, northerly mi- gration (Steele 1991. Lyons 1995). This study was based on commercial production from Apalach- icola Bay. This area is described in Ingle ( 1951 ). Ingle and Daw- son ( 1953). and Livingston and Joyce ( 1977). It was conducted to determine the size and sex of hard blue crabs processed by com- mercial operations. The average yearly volume of hard blue crabs landed in the Apalachicola Bay area (Franklin County) during the period 1974— 1976 was more than 1 ,000,()()0 pounds, according to Landrum and Prochaska (1980). They describe a declining trend in pounds pro- duced as the result of a decrease in effort that had prevailed since 1964. but increasing dockside prices compensated monetarily for the voluine decline. The increase in dockside price and the con- sequent value of the fishery also offset the decrease in the number of fishermen, firms, and traps. The decline in production continued after 1976. During a re- cent span of years (1990-1995). the average pounds landed was 258,000. During the last year of that period (1995). only 122.000 pounds were produced (Kennedy 1995). Conversations with crab house owners indicated that the de- crease in fishing effort is largely due to a bottleneck created by a decrease in labor. Funding by social programs competes for po- tential fishermen and employees who might seek work as crab pickers, according to industry representatives (personal commu- nications 1981 and 1995). PROCEDURE Crabs caught in Apalachicola Bay and processed locally were sampled intermittently for a 2-y period (June 1979 to May 1981). Samples were measured after crabs had been cooked. Each sample consisted of a basket approximately 45 cm\ Two samples were chosen randomly from the picking room of the crab house on each sampling date. Carapace widths between the extremities of the lateral spines of all crabs within the two baskets were then mea- sured. Data were recorded in size classes of a 4-mm range, from 80 to 200 mm. and divided into four categories. These groups were males, mature females, gravid females, and immature females. Crabs smaller than 80 mm were measured and recorded but were not included in any analyses because the numbers were insignifi- cant. Results from samples were totaled for each group and plotted on graphs by size frequencies. Range, mode, and mean of cara- pace width for each category were also plotted. Several factors affected the size of crabs brought into the pro- cessing plant. The commonly used 3.25-cm-pore-size mesh chicken wire of which the traps were constructed prevented most crabs smaller than the mesh size from being captured. This is not critical because crabs of this size are not of commercial value. Peelers may have been removed for soft-shell production and not sold to the picking plant. The number of crabbers, traps per boat, and location of harvesting varied seasonally depending on the availability of crabs. The assumption is made that local fishermen place their traps in areas where prospects are high for catching desirable sizes of animals in commercial quantities. RESULTS The first sample was obtained in June 1979; the last was ob- tained in May 1981. During December 1980 and February and March 1981 , very few crabs were harvested. No data are available for these months. The total number of crabs and the number of sainples taken (N) are indicated for each month in Table 1 , which also shows the mean size of all categories combined for each month. The number of crabs per basket examined depended on the size of the animals and to some extent on the random alignment of the animals within the basket. No attempt was made to "pack'" the crabs. The basket size containing the sample was always the same, but more crabs of smaller size were required to fill it than crabs of larger size. These data indicate that essentially the same size crab is processed year-round. Only 20 mm separated the low mean for June 1980 and the high mean for January 1981. The percentage of each category caught each month is shown in 763 764 Ingle and Ingle TABLE 1. Number of crabs and samples taken and mean size of all categories combined for each month. Month Total Mean I mm) June 1979 May 1980 June 1980 July 1980 August 1980 September 1980 October 1980 November 1980 January 1981 Apnl 1981 May 1981 625 1,309 862 1,122 631 964 817 629 216 1,234 886 135 139 132 132 136 137 146 148 152 141 142 " n. number of baskets. Figure 1. The largest percentage of male crabs by month was caught in June 1980. This percentage decreased during winter months and increased during the rest of the year. The percentage of mature females was highest in the winter and lowest during the summer. This seasonal change in relative abundance is compatible with previous observations in other blue crab studies. Female crabs move to deeper waters in warmer months for hatching (Cargo 1958. Fiedler 1930, Tagatz 1968). From the data available, it appears that two peaks in the per- centage of gravid females in the total population occur: one in August 1980 (27%) and one in April 1981 (25%). In general, gravid females are only a small percentage of the commercial catch, so the data above are worth noting. Gravid females were not seen in any samples in November 1980 or January 1981. No samples were taken in February or March 1981, so information is not available on the time that spawning began. However, by April, gravid females were fairly abundant in the commercial catch. Immature females comprised only a small percentage of the crabs harvested each month. However, small numbers were present in every month sampled. The range, mode, and mean of carapace width of all categories 80 70 CO 43 60 2 ^ 50 O d 40 Z ^ 30 O 20- o^ 10- I I 1 1 1 I I 1 1 1 1 — 6-79 5-80 6-80 7-80 8-80 9-80 10-80 11- 1-81 4-81 5-81 Month Figure I. Percentage of total (Tot. I number (No.) of crabs by category for each month. Male ( — ), mature female ( ), gravid female ( ), immature female (■-•-■). 200-1 1 ■ r 1 —. 180- 1 1 T E .§. 160- S 140- : -^ "^ Carapace 00 o ro o o o 6-79 5-80 6-80 7-80 8-80 9-80 10-80 11- 1-81 4-81 5-81 Month Figure 2. Range, mode, and mean of carapace width for males. Mode ( — ), mean ( ). .§. 180- • T "5 160- g g 140- CD §■120- CD O lOOJ ■ 6-79 5-80 6-80 7-80 8-80 9-80 10-80 11- 1-81 4-81 5-81 Month Figure 3. Range, mode, and mean of carapace width for mature fe- males. Mode ( — ), mean ( ). ^-^ 190 E E. 170 sz ■o IbO g 0 130 o CO 110 CD O 90. — I 1 r- — I 1 1 1 1 I 1 1^ 6-79 5-80 6-80 7-80 8-80 9-80 10-80 11- 1-81 4-81 5-81 Month Figure 4. Range, mode, and mean of carapace width for gravid fe- males. Mode ( — ), mean ( ). 160' S- 140. ;g 120. g O 100' o CD a. CD 80. — I 1 1 I 1 1 1 1 I I P- (0 6-79 5-80 6-80 7-80 8-80 9-80 10-80 11- 1-81 4-81 5-81 " Month Figure 5. Range, mode, and mean of carapace width for immature females. Mode ( — ), mean ( ). Notes on Blue Crab Fishery 765 TABLE 2 TABLE 2. Summary: carapace width (range , mode, and mean) of al continued categories by month. Carapace width (mml Carapace width (mm) Total No. of Individuals Total No of Sex Range Mode Mean Sex Individuals Range Mode Mean June 1979 mmature female 108 88-153 123 123 Male 446 98-173 128 128 Total 1,234 n = 8 Mature female 167 118-178 133 144 M ly 1981 Gravid female 10 138-183 138 154 Male 595 83-183 133 139 Immature female t 83-103 103 93 Vlature female 228 123-188 158 153 Total 625 n" = 4 Gravid female 27 138-183 158 160 May 1980 mmature female 36 83-143 128 119 Male 786 311 83-188 123-193 128 148 136 152 Total 886 n = 6 Mature lemale " n , number of baskets. Gravid female 86 123-188 153 153 Immature female 126 83-148 118 117 Total 1.309 n = 8 June 1980 90-1 Male 661 83-173 118 129 Mature female 164 123-183 143 145 SO- A Gravid female 4 148-173 173 162 / \ Immature female 33 93-138 118 116 TO- / \ Total 862 n = 4 60- / \ July 1980 >< / \ Male 842 83-183 128 128 o c 50- / ^^ Mature female 192 108-188 138 142 or / \ Gravid female 81 113-178 153 144 40- / \ Immature female 7 93-128 113 112 9? 30- / \ Total 1,122 n = 6 / * \ August 1980 20- / / ^ \ * ' > Male 285 98-188 118 131 Mature female 175 108-193 143 146 10- J /' ^^~\"^-^ Gravid female 169 9 98-178 108-128 143 118 136 116 ^— ' ..■-.^->-r^":..-.- Immature female 8 0 90 100 110 120 130 140 150 160 170 180 190 200 Total September 1980 Male 631 n = 4 Carapace witdth (mm) 659 98-173 133 134 Figure 6. Size frequencies for all categories during June 1979. n = 8. Mature female 147 118-183 148 146 Male ( — ), mature female ( ), gravid female ( ), immature Gravid female 111 123-178 148 150 female ( - - ). Immature female 47 108-143 118 121 Total 964 n = 6 October 1980 Male 535 108-178 138 143 90-1 Mature female 261 123-188 155 152 Av Gravid female 7 138-168 148 151 80- / \^-\ Immature female 14 113-128 113 118 / ^'x Total 817 n = 6 70- K / \ November 1980 60- / ^ \ Male 212 98-183 138 141 >, j \ Mature female 401 118-193 158 153 o c 50- / \ Gravid female 0 1 r- \ Immature female 16 88-133 123 121 (T 40- \ ' ^^\_ Total 629 n = 4 2 LL 30- / r-A / V January 1981 // ^^. ,'-' A V, Male 58 118-198 148 152 20- Mature female 157 128-173 148 152 // ^' .••■■ '•.. Nf Gravid female 0 10- /^ • '^ — •< >s; Immature lemale 1 216 111-115 n 113 — 2 113 =^^='^7~v "v- ■■■■■■O^"- Total 8 0 90 100 110 120 130 140 150 160 170 180 190 200 Apnl 1981 Male 568 88-188 118 133 Carapace wi(dth (mm) Mature female 255 118-183 158 154 Figure 7. Size frequencies for all categories during May 1980. n = 8. Gravid female 303 113-188 148 149 Male (- -), mature female ( ), gravid female ( ), immature 766 Ingle and Ingle iiOn 60-1 80 Figure 8. Male (— ) female (- - . -I 1 1 'I 'I ' I ' I 1 90 100 110 120 130 140 150 160 170 180 190 200 Carapace width (mm) Size frequencies for all categories during June 1980. n = 4. , mature female ( ), gravid female ( ). immature are shown in Figures 2-5 and Table 2. This table also shows the total numbers of crabs measured. In general, males were smaller than mature and gravid females. Immature females had the small- est carapace width of all categories. The largest range in carapace width occurred in males (83-198 mm); the smallest range occurred in immature females (83-153 80 100 110 120 130 140 150 160 170 Carapace wi(dth (mm) T 180 190 200 Figure 9. Size frequencies for all categories during .July 1980. n = 6. Male ( — ), mature female ( ), gravid female ( ), immature female (• - • - ■). 80 90 100 110 120 130 140 150 160 Carapace width (mm) Figure 10. Size frequencies for all categories during August 1980. n = 4. Male ( — (, mature female ( ), gravid female ( ), imma- ture female (•-■-■). T 150 T 160 T" 170 190 200 130 140 Carapace width (mm) Figure II. Size frequencies for all categories during September 1980. n = 6. Male ( — ), mature female ( ), gravid female ( ), immature female (•-■-■). 90- 80- 70- o c 0) cr 2^ 40 60- 50 30 20- 10- 0 -r- 80 90 — r 100 I r 150 160 — T- 170 I I I ISO 190 200 110 120 130 140 Carapace width (mm) Figure 12. Size frequencies for all categories during October 1980. n = 6. Male ( — ), mature female ( ), gravid female ( ), im- mature female (■-•-•). Notes on Blue Crab Fishery 767 mm). Mature females ranged from 108 to 193 mm. Figures 6-16 show the same information as Figures 2-5. plotted by month rather than categorized by sex for all four groups. DISCUSSION Laughlm ( 1979) found that larger crabs were most abundant m spring and summer in the Apalachicola Estuary. High temperature and salinity appeared conducive to the occurrence of larger crabs. Our data indicate a slight tendency toward more small crabs during summer than winter months. However, the size differences are small (Table 2). It appears that essentially the same mean size crab is caught year-round. Male crabs, with a smaller mean size than mature females, are relatively less abundant in winter (November 1980 and January 1981 ). Cooler temperatures apparently depress the development of eggs; therefore, during lower temperatures, many individuals that would be in the gravid category (under warmer conditions) are lumped with mature individuals when eggs were not visible. Fe- males (mature) with a larger mean size exerted a greater influence on the overall mean size of crabs in winter samples because of the relative decline in the numbers of males of lower mean size. Because spawning occurs during the warmest months of the year, the number of gravid females would be expected to increase during these months, with a resultant decrease in the relative num- ber of mature females. Our data reflect this: mature females were proportionately greater in winter than in summer months. Putatively, it is illegal to harvest gravid females in Florida. However, egg-bearing crabs are processed. Less than 10% of the catch is composed of these females. Because females spawn off- shore and most traps are placed inside the protection of St. George Island, there is a built-in bias in our data. Because gravid females migrate offshore to spawn, their relative abundance in the estuary declines. Even if the Apalachicola region is not the main spawning ground for the blue crab on the upper western coast of peninsu- lar Florida, as proposed by Oesterling et al. (1982). the area (Apalachee Bay) does maintain a large population of blue crabs. The population is of such a large size that Laughlin ( 1979) felt that cannibalism, by larger crabs, might play a role in the observed population distribution. 70 n 90- 80- 70- /• ^ 60- § 50- o- Q^ 40- LL 1 \ 1 1 1 1 f 1 1 1 f 1 / \ 30- / K \ \ \ 20- A \ 1 \ \ 10- ^./ / .^r- 1 ^"^ 0 ■ r" ' t^ 1 80 90 100 110 120 130 140 150 160 170 180 190 200 Carapace wi(jth (mm) Figure 13. Size frequencies for all categories during November 1980 n = 4. Male ( — ), mature female ( ), gravid female ( ), im- mature female (•-■-■). I I I 1 1 1 I I 1 1 90 100 110 120 130 140 150 160 170 180 190 200 Carapace witjth (mm) Figure 14. Size frequencies for all categories during January 1981. n = 2. Male ( — ), mature female ( ), gravid female ( ) [none]), immature female (■--•). It was putatively unlawful at the time of this study for any person to possess for sale blue crabs measuring less than 5 inches (126.5 mm) between the points of lateral spines in an amount greater than 10% of the total number of blue crabs in such person's possession. However, if the fisherman possessed a special permit for peelers or the bait trade, the law did not hold. A special permit was easily obtained from the Department of Natural Resources. Crabs less than 126.5 mm could be sold as peeler crabs, which bring a higher price. It should be noted that at the time of this study, regulations were basically ignored when they concerned crab harvests in the area. The few crabs smaller than 126.5 mm that entered the process- ing house were predominantly immature females, which com- prised 10% or less of the total catch each month. Mature and gravid females were seldom less than the legal size. On three occasions (June 1980, August 1980. and April 1981). the male carapace width mode was less than the prescribed length. The mean of these crabs on those dates, however, was larger than 126.5 mm. All of the data are presumed to be affected by the proclivity of males and females to favor different habitats during certain periods of their life cycles. Males tend to occupy shallows, whereas fe- males are known to move into deeper water, especially at spawn- ing time. The data presented here must be judged against a back- ground of highly variable physiologic functions, some of which result from unstable environmental conditions. Tagatz (1968) re- ported on studies made in the St. Johns River, its estuary, and offshore ocean waters. He found that after reaching maturity, fe- males were impregnated. Truitt ( 1939) found that multiple spawn- ings can be expected from one impregnation. Although the transfer of sperm from male to female usually occurs in less saline waters, after spawning, the females migrate to saltier, often offshore, wa- ters when hatching occurs. This movement has also been noted by Fiedler ( 1930) and Cargo ( 1958). In the St. Johns River area, the female migration offshore was most pronounced in the spring and fall (Tagatz 1968). Tagatz concluded that larval development took place mainly in the ocean and that the final metamorphosis from magalops to first crab stage takes place most frequently there. As in the St. Johns' area. Van Engel ( 1958) reported that when females migrated to saltier waters (in Chesapeake Bay), males generally tended to remain in areas further upstream. Postlarval instars were estimated to be 20 for males and 1 8 for 768 Ingle and Ingle 90-1 80 90 100 110 120 130 140 150 160 170 180 190 200 Carapace width (mm) Figure 15. Size frequencies for all categories during April 1981. n = 8. Male ( — ), mature female ( ), gravid female ( ), imma- ture female (■-■-•). females (Newcombe et al. 1949). They also concluded that envi- ronmental conditions, especially salinity, intluence the percent increase in size per molt. Most of the crabs of commercial size caught during most of the year in the lower St. Johns River were females (Tagatz 1968). Nearly all crabs caught in the ocean were females, except in the fall. Tagatz also found that the population of males and females that matured at a small size was larger in salt water than in fresh 80 90 100 110 120 130 140 150 160 170 180 190 200 Carapace width (mm) Figure 16. Size frequencies for all categories during May 1981. n = 6. Male ( — ), mature female ( ), gravid female ( ), immature female ( - - • ). water. The smallest mature females measured in the St. Johns River were 90 mm. The largest immature female was 177 mm. Because crab production varied, primarily because of weather, the number of samples per month in our study was not uniform. For instance, only two samples were checked in January 1981, whereas in two months. May 1980 and April 1981, data were derived from eight samples each month. ACKNOWLEDGMENT The authors acknowledge gratefully the assistance of Michael Landrum in data acquisition. Bames, E. W. 1904. Preliminary Inquiry into the Natural History of the Paddler Crab {Callinecleslhastatiis) With Remarks on the Soft-Shell Crab Industry in Rhode Island. Rhode Island Commissioners of Inland Fishenes. 34th Annual Report, pp. 69-73. Cargo, D. G. 1958. The migration of adult female blue crabs, Callinecles sapitlus Ralhbun, in Chincoteaguc Bay and adjacent waters. J. Mar. Res. 16:180-191. Churchill, E. P., Jr. 1921 Life history of the blue crab. Bull. U.S. Bur Fish, for 1917-1918. 36:91-128. Darnell, R. M. 1959. Studies of the life history of the blue crab {Calli- nectes sapidus Rathbun) in Louisiana waters. Trans. Am. Fish. Soc. 88:294-304. Fiedler, R. H. 1930. Solving Ihc question of crab migrations. Fishing Gazette 47:\S-2\. Fischler, K. T. & C. H. Walburg. 1462. Blue crab movement in coastal South Carolina, 1958-59. Trans. Am. Fish. Soc. 91:275-278. Gulf States Marine Fisheries Commission. 1970. Blue Crab Fishery of the Gulf of Mexico, United States: a Regional Management Program. Gulf States Marine Fisheries Commission, Ocean Springs, MS. Hay, W. P. 1905. The life history of the blue crab {sapidus) U.S. Bur. Fish. Rept.for 1904. .^95-413. Ingle, R- M. 1951 . Spawning and setting of oysters in relation to seasonal environmental changes. Bull Mon. Sci. Gulf Caribh. 1:111-135 Ingle, R. M. & C. Dawson. 1953. A Survey of Apalachicola Bay. Tech. Ser. 10. Florida State Board of Conservation, Tallahassee, FL. Judy, M. H. & D. L. Dudley. 1970. Movements of tagged blue crabs in North Carolina waters. Comm. Fish. Rev. 32:29-35. Kennedy, S. 1995. Landings Report 1995. Institute of Manne Research, Department of Environmental Protection, St. Petersburg, FL. Landnim, P. D. & F. J. Prochaska. 1980. The Florida Commercial Blue LITERATURE CITED Crab Industry: Landings, Pnces and Resource Productivity. Fla. Sea Grant College Repl. No. 34. Laughlin, R. A. 1979. Trophic Ecology and Population Distribution of the Blue Crab, C. sapidus Rathbun, in the Apalachicola Estuary (North Florida. U.S.A.). Dissertation. Florida State University. Livingston, R. J. & E. A. Joyce, (eds.). 1977. Proceedings of the Con- ference on the Apalachicola Drainage System. Flonda Department of Natural Resources Mar. Res. Publ. 26. Lyons, W, 1W5. Mar. Res. Inst., Department of Environmental Protec- tion, St. Petersburg, FL. Newcombe. C. L. , M. D. Sandoz & R. Rogers-Talbert. 1949. Differential growth and characteristics of the blue crab, Callinecles sapidus Rath- bun. J. t'.v/). Zool. 110:113-152. Oesterling, M J. 1976. Reproduction, Growth, and Migration of Blue Crabs Along Flonda's Gulf Coast. Fla. Sea Grant publ. 76-003. Oesterling, M. J. & C. A. Adams. 1982. Migration of blue crabs along Flonda's gulf coast, pp. 37-57. In: H. M. Perry and W. A. Van Engel (eds.). Proc Blue Crab Colloquium, Publ. F. Gulf States Marine Fish- eries Commission, Ocean Swings, MS. OtwelfS. W, J. C. Cato&F. G. Halusky. 1980. Development of a Soft Crab Fishery in Flonda. Fla. Sea Grant Publ. Rept. No. 31. Steele. P. 1991 Population dynamics and migration of the blue crab. Callinectes sapidus (Rathbun), in the Eastern Gulf of Mexico. Proc. 40th Ann. GulfCarihh. Fish. Inst. 1987:241-244. Tagatz, M. E. 1968. Biology of the blue crab, C. sapidus Rathbun, in the St. Johns River, Florida. U.S. Fish Wildlife Ser. Fish Bull. 67:17-33. Truitt, R. V. 1939. Our water resources and their conservation. Chesa- peake Biol. Lab. Solomons, Md. 27:l03p. Van Engel, W. A. 1958. The blue crab and its fishery In Chesapeake Bay. Joiirihil of Shellfish Research. VoL 15, No. 3. 769-775. 19%. A METHOD FOR QUANTITATIVELY SAMPLING NEKTON ON INTERTIDAL OYSTER REEFS ELIZABETH WENNER, H. RANDALL BEATTY, AND LOREN COEN Marine Resources Research Institute Charleston. South Carolina 29422 ABSTRACT We developed a sampling methodology using a 24-m^ lift net to quantitatively sample intertidal oyster reefs as a part of a long-term study of their functional ecology. The method involved surrounding an area of oyster reef with a buried net at low tide, allowing the water level to rise, raising the net at high tide to trap motile organisms, allowing the water to recede, and collecting the entrapped nekton. Natural and artificially constructed reefs were sampled, and efficiency (mark-recapture) studies were performed to evaluate the method. The advantages of this method are; ( 1 ) the habitat in the area to be sampled receives minimal damage; (2) the size and shape of the net system are flexible and can be adapted to fit a variety of habitats; ii) no permanent structures, other than a shallow perimeter trench, are present to act as attractants; and (4) it is relatively inexpensive to purchase and maintain gear. One disadvantage to the method is that it is very labor intensive, typically using three to five people This method proved more efficient on natural reefs than artificial reefs, and the return rate was slightly better for t'undiilus heterodilus than for Palaemonetes spp. Seventeen decapod and 24 fish taxa were collected from initial spring, summer, and fall 1995 sampling. KEY WORDS: Lift net. sampling techniques, intertidal oyster reefs, fishes, decapod crustaceans, habitat complexity INTRODUCTION Oyster reefs are a conspicuous feature of the intertidal zone in most estuaries in the southeastern United States (portions of North Carolina, South Carolina, Georgia and portions of northeast Flor- ida). By forming intensive biogenic intertidal reefs, often adjacent to emergent marsh vegetation, Crassostrea virginica provides the only three-dimensional structural relief in an otherwise unvege- tated, soft-bottom, benthic habitat. In areas otherwise devoid of naturally occurring hard substrate, the many crevices and expan- sive surface area found within an oyster reef provide a refuge and attachment for numerous small invertebrates (e.g.. Dame 1979, Bahr 1974. Klemanowicz 1985. Powell 1994). Although intertidal oyster reefs are prominent in the region, information is generally lacking on the importance of these reefs as habitat for juvenile and adult fishes, crabs, and shrimp, which move on and off the reefs with the tide. Anecdotal information for South Carolina suggests that fishes such as bay anchovy Amhou mitchdU. silversides Menidia spp., and killifishes Fundidus spp. are attracted to oyster reefs because of their complex three- dimensional structure, which provides them with a refuge from fish predators (e.g., sciaenids and paralichthid flounders). Large predators including spotted seatrout Cynoscion nehidosiis. red drum Scicienops oceUatus, summer flounder Parcdichlhys denla- lu.'i. and sheepshead Archosargus probatocephalus migrate onto oyster reefs on flood tides to consume small crabs, shrimps, and fishes that reside in and around the reef structure (Cocn et al. 1996). Oyster reefs in high-salinity waters are also an important habitat for juveniles of several important fish species such as sheepshead A. probatocephalus, gag grouper Mycteroperca mi- crolepis. and snapper Liiljiiniis spp. , as well as stone crab Menippe mercenuria and blue crab Callinectes sapidus (Cain and Dean 1976, Grant and McDonald 1979, Reiss and Dean 1981, Crabtree and Dean 1982, Wilson et al. 1982. Kleypas and Dean 1983, Wenner and Stokes 1984). Quantifying the use of oyster reefs by various life history stages of motile fauna has been limited by sample gear. To our knowl- edge, only Bahr (1974), Crabtree and Dean (1982), and Powell (1994) have previously attempted to quantify transient nekton as- sociated with intertidal or shallow subtidal oyster reefs. Powell (1994) conducted a visual census by diving on reefs, where he observed numerous pinfish Lagodon rhomboides and several sheepshead A . probatocephalus at high tide. Poor visibility (< 10- 15 cm) precludes comprehensive visual censuses on most south- eastern oyster reefs, and disturbance by divers in these shallow areas likely biases results. Open-topped traps (Crabtree and Dean 1982) are biased for the collection of small nekton and do not provide data for the quantification of densities within a portion of the reef habitat. Conventional methods for sampling soft-bottom habitats, such as trawls and seines, cannot be used on intertidal oyster reefs because of hangs and tears from oyster clusters. Other devices such as drop samplers (Wenner and Beatty 1992) can be deployed in hard-bottom habitats, but they sample a relatively small area (<10 m") and may not seal properly along the bottom when placed over dense clusters of living oysters. Poisons, pri- marily rotenone, have been used to sample areas with oyster hab- itat (Weinstein 1979), but the broad spatial effect of the treatment precludes the determination of specific habitat use. This article describes a modification of the lift net (Rozas 1992) and fiume weir (Kneib 1991 ) used to quantitatively sample nekton in emergent vegetation such as intertidal marshes. Our net system is designed to completely surround a defined (e.g., 24-m~) area of intertidal oyster reef, with minimal disturbance during deploy- ment. It IS being used to quantitatively sample nektonic species as part of a multiyear interdisciplinary study that determines the func- tional role of intertidal oyster reefs in southeastern estuaries (Coen ct al. 1996). Although the purpose of this article is to describe the lift net system and its efficiency, we provide data on species com- position and the abundance of fishes and decapod crustaceans collected by the gear to emphasize its effectiveness and versatility. MATERIALS AND METHODS Study Sites and Reef Fabrication Two study areas with similar salinity regimes, bed grades, base sediments, wave disturbance, adjacent oyster communities, and elevation above mean high water were selected near Charles- ton Harbor, SC (Fig. 1 ). One area was located near Toler's Cove Marina (henceforth referred to as the Toler's site), a moderate- sized marina with approximately 138 slips, located within a small tidal creek (depth <3 m). Oyster reefs in this area are bordered by an extensive Spartina aherniflora salt marsh. A previous study by 769 770 Wenner et al. Figure 1. Study area located east of Charleston Harbor, SC. The Toler's Cove site (located around 32°46.27' N latitude and 79°5I.16' W longitude) is a disturbed site because of marina activity, whereas the Inlet Creek site (located around 32°48.02' N Latitude and 79°49.50' W longitude) has no adjacent development. Van Dolah et al. (1992) measured the levels of contaminants, oyster growth and general health and density of spat settlement at this site, which is closed to shellfish harvesting. The second area selected for study was located in the upper reaches of Inlet Creek (henceforth referred to as the Inlet site), a relatively pristine site, with broad expanses of reefs under private lease, a large buffer zone (>3(X) m) of S. alternitlom. and relatively little adjacent development. Both sites are dominated by fine sediments, often >75% silt/clay with little or no sand. Water quality variables of temperature, salinity, and dissolved oxygen, as determined from deployed Hydrolab Datasonde 3S®, were similar between sites (Coen et al. 1996). Both sites are subject to semi-diurnal tides, with a mean range of 1.5-2 m. Three natural oyster reefs and three artificially constructed reefs (henceforth referred to as experimental) were sampled at each site (Fig. 2). Experimental reefs were constructed of 0.46- x 0.31- X 0. ll-m perforated plastic trays lined with a 1.3-mm fi- berglass mesh screen and filled with oyster shell. Trays were filled with shell (~8 kg each) to a standard height of ~0.11 m. All oysters and shell were removed from an area equivalent to reef size (roughly. 8.2 x 2.9 m) before fabrication of the experimental reefs in October 1994. Each experimental reef measured 2.92 x 8.17 m (23.86 m') and consisted of 26 rows of six trays placed end to end. Experimental reefs were designed to approximate the size of a natural oyster reef and to be sufficiently large so as to avoid repeated sampling and disturbance of areas sampled for another component of the overall study dealing with resident reef species (Coen et al. 1996). In close proximity to each experimental reef, an adjacent area of natural oyster reef, 24 m^ (3 x 8 m) with approximately the same elevation and configuration and with live oysters and shell, was marked for sampling and is henceforth referred to as a natural reef. The paired natural and experimental reefs combined to make up more than half of the total reef area or "mound"" on which they were located. Nel Design Our sampling procedure was adapted from methods used pre- viously to sample intertidal vegetated habitat (Mclvor and Odum 1986. Kneib 1991.Rozas 1992. and Wenner and Beatty 1992) that involved isolating a discrete area of habitat with a net and extract- ing the animals trapped within the area. Samples were collected by surrounding an experimental or natural reef with a lift net at low tide when the reef was exposed (Fig. 3). allowing time for the tide to rise, raising the lift net at high slack water during daylight hours, allowing the water to recede, and collecting specimens trapped in the net. Six months after reef construction and a few weeks before sampling, site preparation was completed. This involved driving 12 1.5-m sections of I2.7-mm concrete reinforcing rods (rebar) into the substrate surrounding each rectangular reef so that one was in each comer, three were spaced evenly along each long side, and one was placed in the center of each end. These were to mark placement and to support the net poles. A shallow trench was then excavated outside of the rebar, and a stainless steel cable, 4.8 mm in diameter, was secured in the bottom of the trench with a J-shaped 9.5-mm rebar. The purpose of the cable was to prevent the lifting of the bottom of the net during the net-raising proce- dure. The site perimeter was protected by placing 0.75- x 2.2-m removable plywood walkways around it during construction. Sampling apparatus was set up on the low tide preceding the high tide during which sampling would take place. Before the nets were rigged, the site perimeter was again protected by the place- ment of the plywood walkways around it. Walkways were re- Nekton Sampling Method 771 Figure 2. Conceptual layout of a reef site with three pairs of reefs, one natural reef and one experimental reef in each pair. moved once the site was ready, so as to not interfere with ambient flow around the reef. An aluminum pole —2.2 m long was placed over each of the rebar supports and driven into the substrate. The length of the poles was determined with data on tide height, so that the tops of the poles, when in place, were —20-30 cm above high water. The lift net was constructed of 3.2-mm-pore-size mesh nettmg that was 2.44 m wide. The net was long enough to enclose the site and overlap ~2 m so that a secure seal could be achieved (Fig. 3A). A sleeve was sewn into the top and bottom of the net. with a line placed in the top sleeve for support and a 6.4-mm cham fed through the bottom sleeve to help weight the bottom and achieve a tight seal. After the net encircled the site, the net was secured to the cable with cable ties —10 cm above the chain to avoid the lifting of the bottom of the net when the net was raised. The lift net was then folded into the shallow trench (Fig. 3B and C) so that the top of the net was approximately even with the top of the trench; the net was then covered with sediment (pluff mud) to help conceal it and hold it down over a tidal cycle. Pit trap Oysters Figure 3. Reef mound containing a pair of reefs: an artificial reef (A) (with the lift net up) and a natural reef (B) (with the lift net down). Before sampling, the net is folded over the lead line and cable in a shallow trench (C) and covered with sediment. The pull line is at- tached to the top of the net and threaded through an eye bolt in the top of the pole. Five 8.0-mm polypropylene lines were attached to the net in order to raise it from a remote station —15 m away, thereby lessening the degree of disturbance to the sample site before the net was raised (Fig. 3). These lines were attached to rings on the head rope and were threaded through the eye bolt in the top of each pole. The lines at each end were threaded through a ring and eyebolt at each end pole. A pit trap was positioned inside the area sampled by the net at the lowest elevation into which organisms could accumulate as the tide receded. The pit traps were constructed by excavating a hole -30 cm in diameter and -45 cm deep and inserting a polyvinyl chloride bucket with a removable 3.2-mm-pore-size mesh basket. Each pit was covered with a weighted top as the tide was rising to prevent it from biasing the sample in any way. The pit-trap covers were removed after the net was raised. The time of high slack tide was estimated from NOAA tide tables, and the actual time was determined by visually monitoring water current speed at the site. Although each area contained three pairs of reefs (a natural reef and an experimental reef in each pair), only two pairs could be sampled simultaneously because of man- power limitations. Lift nets were raised simultaneously by two teams of four people, who pulled on the lifting lines from the remote stations. Lifting the nets took less than 6 sec. After the nets were raised, the lift lines were tied to the anchor post and the field crew proceeded to secure the top line of the net to the net poles so that the net would not be pulled down as the tide receded. After the water level dropped and the reefs were exposed, the 772 Wenner et al. plywood walkways were again put in pluL-c. The lift net was low- ered to permit two teams of two individuals to move along the walkways in opposite directions and collect all animals trapped at the base of the net. Animals were also retrieved from the pit trap by removing the insert. All specimens were preserved in 10% buffered formalin. In order to minimize disturbance to the reefs, only those animals observed from the perimeter were collected from interior sections of each reef. To date, we have successfully used this technique to sample 1 2 reefs: 3 paired reefs at the Toler's site and 3 paired reefs at the Inlet Creek site, during daylight hours in May, July, and October 1995 (n = 6 reefs/site per date). All specimens collected were returned to the laboratory, iden- tified to species, and enumerated, with the occasional exception of grass shrimp Palaemonetes spp.; P. puf>io and P. vulgaris were sometimes encountered in numbers large enough (> 1,000) to ne- cessitate subsamplmg. Our subsampling procedure consisted of randomly selecting a subsample that was 10% by weight of the total sample of all Palciemoneles spp. and then identifying and sorting the sample by species. Each species was then enumerated and weighed, and the ratios of number/weight in the subsample were used to estimate total number and weight for each species from the total weight of the group. Capture Efficiency Efficiency tests were conducted at two of the paired experi- mental and natural reef sites in Inlet Creek on three consecutive days in March 1995. After nets were raised at high slack tide as described above, 50 specimens of mummichog Fundulus helero- clitiis. with clipped anal fins, and 50 grass shrimp Palciemoneles spp. stained with alcian blue iCoen et al. 1982) were released into the sampling area. These species were chosen because they were readily available at the sampling sites and are not demersal resi- dents of oyster reefs. Five paired replicate tests were conducted, and the mean return rates were used to represent efficiency. A two-sample f-test for independent samples was used to de- termine whether the mean number of recaptured individuals dif- fered between natural and experimental reefs. Variances were de- termined to be equal by use of Levene's test (SPSS. Inc. 1993). Chi-square analysis of pooled frequencies was used to test the null hypothesis that the sample of recaptured individuals fit an ex- pected 1;1 ratio between natural and experimental reefs (Sokal and Rohlf 1981). The significance level in all statistical tests was a = 0.05. RESULTS Of the 17 decapod and 24 fish taxa collected on oyster reefs in May, July, and October 1995, the grass shrimps P. vulgaris and P. pugio were the most abundant (Table I). Overall, these species constituted 70 and 12%. respectively, of the total 27.850 individ- uals captured during the three sample periods. The overall density of P. vulgaris for both sites during the sampling penod was 22 individuals/m", whereas P. pugio occurred at a much lower den sity of 4 individuals/m". Other numerically important species col- lected were the bay anchovy A. nutchilli (5%), the naked goby Gobiosoma bosci (3%), brown shrimp Penaeus aztecus (3%), and white shrimp Penaeus setiferus (2.5%). Grass shrimps, bay an- chovy, and naked goby abundances were consistent during each collection date, whereas brown and white shrimp abundances were high in May and June, respectively. The percentage of individuals recaptured during the capture efficiency study was greatest on the natural reefs, with 68.5% (standard error, SE = 10.1) of the mummichogs recovered and 58% (SE = 11.4) of the grass shrimp recovered. On the experi- mental reefs, 54% (SE = 10.9) of the mummichogs and 43% (SE = 8.5) of the Palaemonetes spp. were recovered. Although no significant difference (two-sample l-lest) was detected between the mean number of individual Fundulus or Palaemonetes recaptured on natural and experimental reefs (Fig. 4), total numbers recov- ered for each species differed significantly from 1:1 for reef treat- ments (pooled x~ = 4.52, p < 0.05 for Fundulus: pooled x' = 5.14. p < 0.05 for Palaemonetes). with more individuals of each species captured on the natural reefs. DISCUSSION The recapture efficiency of the net system we used to sample oyster reefs compared favorably with similar techniques used to sample salt marshes. Rozas (1992) estimated the efficiency of a bottomless lift net to range from 32 to 93%, depending on the species, with the method being less efficient in capturing Palae- monetes (32%) and more efficient (81%) in capturing Gulf killifish (Fundulus grandis). Similarly, Kneib (1991) found that a flume weir recovered F. heteroelitus and P . pugio less efficiently from the marsh surface than it did other species. He (Kneib 1991) con- ducted several efficiency tests where test animals were released into an area that was then sampled three consecutive times. An average of 62% of the released Fundulus and an average of 42- 72%' (depending on size) of the released grass shrimp were recov- ered on the first retrieval efforts of the tests. Various explanations have been given for the efficiency levels of drop nets and flume weirs. Rozas ( 1992) attributed his lowered efficiencies to the escape of organisms through holes in the netting made by blue crabs. We have also found holes (generally <2 cm) around the base of our nets and have frequently seen blue crabs tearing at the net in an attempt to grab organisms caught in the folds of the mesh. Although Rozas (1992) discounted a possible loss of efficiency due to the avoidance of collecting pans by or- ganisms remaining on the marsh, such avoidance may be a sig- nificant factor in our study. In our study, the additional structure created by the experimental trays and complete coverage by oyster shell on the experimental reefs undoubtedly provided more refuges and natural depressions than the patchy shell density of natural reefs, thereby reducing capture efficiency. Also, our reefs were sufficiently large that thorough inspection was not possible with- out walking on the reef, a situation deemed to be an undesirable disturbance. However, small dip nets were used to strain water- filled depressions that formed along the base of the net and trays to recover organisms using them as refuges. Our modification of the lift net system used previously to sam- ple emergent vegetation has many of the same advantages noted by Rozas (1992) for his sampling of intertidal marshes. Because we deployed the nets immediately before each sampling event and took them down at completion, there were no permanent posts or walkways and little habitat modification. A shallow trench to hold the net was the only necessary construction. It was fitted with netting covered by mud so that nekton could move onto the reef without being impeded or scared away by the net. Another major advantage of this technique is its flexibility in size for sampling a variety of intertidal habitats. Rozas (1992) described the use of the bottomless lift net for sampling small (6-m"), discrete salt marsh areas. Our adaptation is designed to Nekton Sampling Method 773 TABLE I. Rank abundance for species collected on intertidal oyster reefs in May, July, and October 1995 (IC = Inlet Creek; IC = Tolers Cove).' IC Control IC Experimental TC Control TC Experimental (No. of (No. of (No. of (No. of (ilrand Abundance Rank ( % Species Name Individuals) Individuals) Individuals) Individuals) Total {%) Abundance) A Ipheus hewriichuelis 1 0 0 1 2 0.01 40 Anchoa mitchilli 619 311 321 129 1 .380 4.96 3 Archosar^us prohalocephahis 1 0 2 3 6 0.02 31 Bairdiellu chrysoiiru 10 6 1 1 18 0.06 24 CiillinecU's \o/)/(/h.v 26 21 14 11 72 0.26 11 Calliueclcs siniitis 3 3 0 0 6 0.02 31 Chasmodes hosqiuiiinis 10 16 62 14 102 0.37 10 Clihanarius vilUilus 0 0 9 0 9 0,03 30 Eucinostomus ar^enteits 60 42 44 88 234 0.84 7 Emiiwslomus sp. 0 0 14 12 26 0.09 20 Eurypanopeus depressus 3 0 1 1 5 0.02 35 Eun'tium limosum 6 10 23 16 55 0.2 14 Evorlhodtis lyricus 0 0 1 0 1 0 41 Fundulus heleroclilus 4 6 5 1 16 0.06 26 Gobionellus holeosoma 21 18 25 42 106 0.38 9 Gobiosoma hosci 84 217 290 304 895 3.21 4 Lagodon rhomboides 7 9 31 10 57 0.2 12 Leiostomus xtinthitrus 4 8 23 9 44 0.16 16 Eitljanus grisetts 4 9 6 15 34 0.12 18 Menidia menidia 12 97 17 10 136 0.49 8 Miigil cephalus 5 7 -) 2 16 0.06 26 Mugil curema 0 3 13 25 41 0.15 17 Myrophis punclatus 0 0 1 0 1 0 41 Opsanus lau 0 0 4 10 14 0.05 28 Orthopristis chrysoplerus 0 0 4 13 17 0.06 25 Palaemonetes pugio 146 237 872 2.142 3.397 12.2 2 Pataemonetes sp. 0 0 0 5 5 0.02 35 Palaemoneles vulgaris 1,699 1,910 8.673 7.101 19.383 69.6 1 Panopeits herbstii 10 6 10 6 32 0.11 19 Panopeus obesus 11 14 15 16 56 0.2 13 Panopeus sp. 9 7 3 3 22 0.08 23 Paralichtlns lethostigma 1 0 0 0 1 0 41 Paralicthvs dcnlatus 4 3 9 7 23 0.08 22 Penaeus aziecus 238 177 332 88 835 3 5 Penaeus duorarum 3 0 1 2 6 0.02 31 Penaeus setiferus 213 77 322 82 694 2.49 6 Sciaenops ocellalus 2 3 2 3 10 0.04 29 Sesarma sp. 0 1 0 0 1 0 41 Syngnathus louismnae 1 1 0 2 4 0.01 38 Uca minax 1 3 0 1 5 0.02 35 ilea pugilalor 6 1 5 13 25 0.09 21 Uca pugiuLX 1 1 4 0 6 0.02 31 Uca sp. 15 9 8 15 47 0.17 15 Total 3.242 3,235 11.170 10.203 27.850 100 Number of individuals is total number of specimens from three experimental or three control reefs at each site sample a larger (24-m") intertidal area, regardless of reef shape. We caution that the expansion of the area sampled might compli- cate raising the net sides by requiring more liftlines and greater deployment time. We cannot attest to the use of this net in marsh habitats but suggest that it would be feasible because of its simi- larity to the net described by Rozas (1992). Although Kncib (1991) noted that the use of a Hume weir in habitats with little or no emergent structure could bias results, our technique likely avoided such bias for several reasons. We did not install a permanent boardwalk around the netting, thereby avoid- ing shadows, which may attract fish. Removable plywood walk- ways around each reef minimized disturbance during sampling. Oyster reefs have a vertical structure ( 15-25 cm) that far exceeds that of the buried net within its trench, so the avoidance or attrac- tion of nekton to the net should be minimal. Escape was mini- mized as the result of the ability to pull the nets upward from a submerged position rather than lowering panels (Kneib 1991 ) or a lead line (Mclvor and Odum 1986, Wenner and Beatty 1992). The use of lift lines from a location away from the net also reduced site disturbance. The height of the net was sufficient to be above the high-water level in estuarine systems with a mean tidal amplitude of -1.5-2.0 m. 774 Wenner et al. Net Efficiency UJ t = 0.9E df=7 ^ , p = 0.37 t= 1 01, p = 0,34 JO df=8 re !5 30 - " > ■a "5 20- O 2 re 10 - u n 100 UJ (/) 80 +1 0) 60 o o o ^'° c re S 20 - n= 4 n =5 n= =5 " n=5 "" - Mummichog Grass Shrimp Mummichog Grass Shrimp Control Experimental Figure 4. Efficiency of recapture for F. heteroclitus and P. piigio released in an area enclosed by a lift net. The mean number of individuals and the mean % recovery from replicate releases on control and experimental reefs are presented. Results of a r-test to determine significant differences among mean numbers of recaptured individuals between control and experimental reefs are shown. The cost of materials was —$1,000 per sampling unit (one reef), including net. chain, line, poles, rebar. and cable. Because nets were deployed and removed in a day and not left in the field for an extended period, they experience little deterioration. Nets were rinsed and dried immediately after sampling to prolong their use. Typically, six paired reefs could he sampled per week. A major disadvantage to this sampling technique is the number of people necessary to set up and deploy each net. We used teams of four individuals per site (two per net) during our study, all of whom could work comfortably out of a single vessel. A pair of nets could be fully prepared in 1-2 h. The logistics of sampling depends largely on the number of replicates in the study. The sampling method described here provides a means for the quantitative assessment of ncktonic species associated with inter- tidal oyster reefs. Previous methods attempting to quantify organ- isms associated with subtidal reefs were not suitable for the cal- culation of density estimates of transient reef species. This method quantitatively samples small reefs (tens of square meters), requires little modification to the reef, is deployed with a minimum dis- turbance of organisms in situ, and is relatively inexpensive to construct and maintain. Information obtained from the sampling of intertidal oyster reefs aids in determining the relative value of various critical habitats in southeastern U.S. estuaries and helps to elucidate how this extensive oyster reef habitat contributes to the broader functioning of estuarine ecosystems. ACKNOWLEDGMENTS This is contribution number 382 from the Marine Resources Research Institute. Numerous individuals were involved in this study, but we are especially grateful for the able assistance of Bruce Slender. David Knott. Mark Thompson. Brett Fallaw. Will Heglcr. Will Shimp. Mike Wert. Chris Graffeo. Will Snead. and Nancy Hadley during sampling and sample processing. This por- tion of the reef ecology project was funded by the S.C. Sea Grant Consortium. The S.C. Marine Recreational Fisheries Stamp Pro- gram, and the S.C. Department of Natural Resources. LITERATURE CITED Bahr. L. M.. Jr. 1974. Aspects of Ihe structure and function of the inter- tidal oyster reef community in Georgia. Ph.D. Dissertalion. University of Georgia. Athens. 149 pp. Cain. R. L. & J. M. Dean. 1976. The annual occurrence, abundance, and diversity of t~ishes in a South Carolina intertidal creek Mcir. Biol 36:369-379. Coen, L. D.. K. L. Heck, Jr. & L. G. Abele. 19X2. Experiments on competition and predation among shrimps of seagrass meadows. £fc)/- ogy 62:1484-1493 Coen. L. D.. D. M. KnoU, E. L. Wenner. N. H. Hadley & A. H Ring- wood. 1996. Intertidal oyster reef studies in South Carolina: design. sampling and experimental focus for evaluatmg habitat value and func- tion. In: M. Luckenbach. R. Mann and J. Wesson (eds). Oyster Reef Restoration Symposium Proceedings. Williamsburg, VA (in press). Crablree. R. E. & J. M. Dean. 1982. The structure of two South Carolina estuarine tide pool fish assemblages. Estuaries 5:2-9. Dame, R. p. 1979. The abundance, diversity and biomass of macro- benthos on North Inlet, South Carolina, intertidal oyster reefs. Proc. Null. Shellfish Assoc. 68:6-10. Grant. J. & J. McDonald. 1979. Desiccation tolerance of Eunpanopeus depressus (Smith) (Decapoda: Xanlhidae) and the exploitation of mac- rohabitat. Estuaries 2:172-177. Nekton Sampling Method 775 Klemanowicz. K. J. 1985, Effects of a mechanical oyster harvester on macrofaunal community structure MS Thesis. The College of Charleston, SC. 102 pp. Kleypas. J. A. & J. M. Dean. 1983 Migration and feeding of the pred- atory fish. Bairdiella chnsura Lacepede. in an intertidal creek, J. Exp. Mar. Biol. Ecol. 72:199-209, Kneib. R. T, 1991 , Flume weir for quantitative collection of nekton from vegetated intenidal habitats. Mar. Ecol. Prog. Ser. 75:29-38, Mclvor, C. C. & W E. Odum, 1986, The Hume net: a quantitative method for sampling fishes and macrocrustaceans on tidal marsh sur- faces. Estuaries 9:219-224. Powell, C. M. 1994. Trophic linkages between intertidal oyster reefs and their adjacent sand flat communities. M.S. Thesis. University of North Carolina, Wilmington. 44 pp. Reiss, R. R. & J. M. Dean. 1981, Temporal variation in the utilization of an intertidal creek by the bay anchovy {Anchou milchilli). Esiiiaries 4:16-23, Rozas, L, P. 1992. Bottomless lift net for quantitatively sampling nekton on intertidal marshes. Mar. Ecol. Prog. Ser. 89:287-292, Sokal.R R,&F, J, Rohlf, 1981 Biometry, W,H Freeman and Co, San Francisco, 859 pp, SPSS, Inc, 1993, SPSS for Windows Base System Users Guide, Release 6,0. SPSS, Inc., Chicago. 828 pp. VanDolah, R. F., M Y Bobo, M, V. Levisen, P. H. Wendt & J. J. Manzi. 1992. Effects of marina proximity on the physiological condi- tion, reproduction, and settlement of oyster populations. J. Shellfish Res. 11:41-48. Weinstein. M. P, 1979. Shallow marsh habitats as primary nurseries for fishes and shellfish. Cape Fear River, N.C. Fish. Bull. 77:339-357. W'enner. E. L. & H. R. Beatty. 1992. Utilization of shallow estuarine habitats in South Carolina, USA, by postlarval and juvenile stages of Penaeus spp. (Decapoda: Penaeidae), J. Crust. Biol. 13:280-295. Wenner, E. L. & A. D. Stokes. 1984. Observations on the fishable pop- ulation of the stone crab Menippe mercenaria (Say) in South Carolina waters. J. Shellfish Res. 4:145-153. Wilson, C, J. M. Dean & R. Radtke. 1982, Age, growth rate and feeding habits of the oyster toadfish, Opsanus tau (Linnaeus), in South Caro- lina. J. Exp. Mar. Biol. Ecol. 62:251-259. 778 Morris and Campbell aerated, running seawater (5 L/min) at ambient water temperature (14— 16°C). Each size category was divided into three feeding treatments: 30 urchins were fed N. hietkecma. 30 were fed Z. marina, and 30 were starved. Within each feeding treatment, ur- chins were randomly assigned to three replicate tanks (10 urchins per tank). Treatments requiring Z. marina and A', luetkeana were provided with enough material so that individual urchins were in close contact with the available food source at all times. Each tank was closely monitored daily for urchin mortality and cleaned as required. After 90 days, the experiment was terminated. The TD and JL of each urchin were measured as described above. Percent change in TD (d) was calculated as d = 100 (T, - To)/Tq. where T, = the mean final TD (in millimeters) for one replicate tank of urchins after 90 days, and Tq = the mean initial TD (in millimeters) for one replicate tank of urchins at 0 days. Percent change in JL (j) was calculated as j = 100 (J, - Jo)/Jo. where J, = the mean final JL (in millimeters) for one replicate tank of urchins after 90 days, and Jy = the mean JL (in millime- ters) of urchins from that particular size group sacrificed before the experiment. The mean and standard error of d and j for the three replicate tanks from each size group per food treatment were cal- culated. The final JL/TD ratio was calculated for each of the wild and treated urchins; means and standard errors for JL/TD were calculated for all urchins in each treatment. Percent mortality (x) was calculated on a weekly basis for urchins within each tank as x = 100 (M,/N||). where M, = the total number of mortalities at a given time, and Ng = the initial number of urchins at 0 days. A two-way analysis of variance ( ANOVA) was used to compare each growth parameter and mortality estimate of urchins between dif- ferent feeding treatments and initial size groups after the data were arcsine transformed. There were significant differences (p < 0.05) in the interaction between size and treatments in the growth pa- rameters so the Tukey test was used for post hoc multiple pairwise mean comparisons (Wilkinson et al. 1992). The power curve of the linear regression form In JL = In a -I- b In TD was used to approximate (with the least-squares method) the relationship be- tween the final JL and TD of individual urchins collected from each of the wild and the three experimental treatments after the data were natural log (In) transformed. Analysis of covariance (ANCOVA) was used to test between the treatment homogeneity of the slope and the elevation coefficients of the regressions (Zar 1984). RESULTS The percent growth of TD and JL declined with the increase in size categories for fed red sea urchins (Fig. lA and B). Growth was greater for urchins fed Nereocyslis than Zostera. and growth was nearly zero for urchins that were starved (Fig. lA and B). There was no significant percent change (Tukey test, p > 0.05) in growth for all sizes of urchins that were starved (Fig. 1). There were significant differences (Tukey test, p < 0.05) in percent change in TD. JL, and the JL/TD ratio between all urchin feed treatments except for the starved and wild sample urchins (Fig. IC). The average mortality per tank was 15% (n = 10). with 95% of these deaths occurring within the first 2 wk of the experiment, probably due to handling effects. There were no significant dif- ferences in mortality (ANOVA. p > 0.05) between feeding treat- ments or size groups. The JL/TD ratio was higher for starved and wild urchins at each size category than for urchins fed Zostera. and it was lowest for those fed Nereocy.Ui.s (Fig. IC). There was no I- w lU 120 H Z 100 -• LU tr (J LU 80 - • Z I- < ^60 "^40 + 20 ■■ 0 LU O CC LU Q. If) -20- I H LU _l <30 + -3 20 •■ < I O 10 I- z LU tr " -10 -L 029 rr 111 02/ H LU "> < 0.2b a n 0,23 III H < T tL 0 21 H o Z III 0.19 _l § < 0,1/ —} I-. ■ ^ NEREOCYSTIS I- STARVED -H • 1 20 -I NEREOCYSTIS ■ ZOSTERA STARVED -H — - ■ ■! I- i ■ ^ 1 15 20 30 10 15 20 25 30 INITIAL TEST DIAMETER (MM) Figure 1. (A) TD growth, (B) JL growth, and (C) Hnal JL/TD ratio in relation to the initial TD of four size groups of red sea urchin juveniles fed .\ereocystis or Zostera or not fed (starved) for 90 days in aquaria. The JL/TD ratios for the "wild sample" are based on red sea urchins at the start of the experiment. Vertical bars are standard errors about the means (dots). significant difference (Tukey test, p > 0.05) in JL/TD ratios be- tween the different size groups of urchins fed Nereocystis (Fig. IC). The ANCOVA confirmed, for JLs adjusted for the covariate TD (slopes were homogeneous), that although there was no dif- Growth of Juvenile Red Urchins Fed Eelgrass 779 TABLE I. Relationship between the final JL (in mm millimeters) and TD by use of the linear equation In JL = In a -t- b In TD for juvenile red sea urchins from the wild and three laboratory treatments." Treatment Regression Coeflicients In a Test Diameter (mm) Min Max Wild sample Starved Zostera Nereocxstis -0.805 a -0.613 a -0.723 b -1.214c 0.796 0.735 0.760 0.866 0.892 0.899 0.802 0.819 80 98 102 105 10 II 16 23 30 30 37 45 " N is the number of urchins in the sample; r" is the coefficient of determination. There was no overall difference (p > 0.051 between four slopes (b) by the use of ANCOVA for the homogeneity of slopes. Elevations (In a) followed by the same letter were not significantly different (p > 0.051. whereas elevations followed by different letters were significantly different (p < 0.05) by the use of pairwise compansons of treatments with ANCOVA. adjusting the In JL values for the covanate In TD. ference (p > 0.051 between wild caught urchins and starved ur- chins, there were significant differences (p < 0.05) in all other combinations between the starved urchins and those ted Zostera and Nereocyslis (Table I). DISCUSSION Nereocystis is clearly a better food for the growth of S. fran- ciscanus juveniles than is Zostera. Although growth in urchin juveniles fed Zostera was almost 50% less than those fed on Nere- ocystis. there are no previous studies showing the capacity of S. franciscanns to digest Zostera. Echinoids are generally incapable of producing the pectinase enzyme required to break down the pectin polysaccharide found in Z. marina (Lawrence 1975. Lowe and Lawrence 1976. Whyte and Englar 1977. J. N. C. Whyte pers. comm.l. The fact that urchins grew after eating Zostera indicated their ability to digest and absorb some nutrients from the plant. The pectin polysaccharide constitutes a large proportion of the cellulose fiber found in Zostera: this would inhibit the diges- tion of the more readily absorbed components. However, chewing by urchins could break down the fibrous component of Zostera to some degree, allowing enzymatic interaction with the digestible components of the plant. Several studies have shown that polysac- charide-degrading bacteria were present in the guts of the sea urchins and have suggested that bacteria may play a role in diges- tion (Guerinot and Patriquin 1981. Lasker and Giese 1954. Yano et al. 1993). Red sea urchins may also have bacteria capable of assisting digestion. Percent change in the TD and the JL of S. franciscanus fed Zostera or Nereocystis declined with an increase in TD size, which according to Lawrence ( 1975). is due to decreased absorption and the assimilation of urchins with increasing age. Also, the increased allocation of nutrients for sexual maturation and gonad develop- ment would reduce the nutrients available for somatic growth (Fuji 1967. Gonzalez et al. 1993). Small, immature gonad development was visible in all S. franciscanus fed Nereocystis and in manv of the urchins fed with Zostera. In contrast to the growth of fed urchins, the ability of 5. franciscanus to maintain the same TD over 90 days of starvation supports the theory that urchins can live in a feast-and-famine environment (Andrew 1989), maintaining their size over periods of starvation and capitalizing on an abun- dance of nutritious food when available by growing rapidly ( Vadas 1977. Larson et al. 1980. Andrew 1986. Lawrence and Lane 1982). The JL/TD ratios of the urchins starved for 90 days in the laboratory were similar to those of the original wild urchins for the same size categories. The large JL/TD ratio suggests that wild urchins were also starved, probably because of the observed gen- eral absence of marine flora attached to the substrate, accompanied by high densities of red urchins (20-30 urchins per m^) at the collection site (A. Campbell unpub. data). High-quality food such as Nereocystis allowed juvenile red sea urchins to rapidly grow their test and jaws at similar relative rates throughout the TD size range observed. Our results on juvenile S. franciscanus generally agree with the interpretation that sea urchins have an adaptive morphological plasticity in which individual urchins may allocate more resources to the food-gathering Aristotle's lantern during food scarcity and allocate more resources to test growth during high-quality food abundance (Ebert 1980. Edwards and Ebert 1991). Although the perennial Zostera is a lower quality food source for S. franciscanus compared with the annual Nereocystis, the presence of Zostera as a drift plant material may be important in maintaining high levels of red sea urchin density when sufficient Nereocystis is unavailable in the coastal waters of Clayoquot Sound. ACKNOWLEDGMENTS We thank D. Brouwer. B. Clapp. J. Clarke. W. Hoyseth. H. Kreiberg. and G. Parker for technical assistance; K. Rajwani for help with the statistical analysis; and D. Bureau. 1. Perry, and J. N. C. Whyte for reviewing earlier drafts of the manuscript. LITERATURE CITED Andrew, N. L. 1986. The interaction between diet and density in influ- encing reproductive output in the echinoid Evechinus chloroticus (Val). J. Exp. Mar. Biol. Ecot. 97:63-79. Andrew . N. L. 1989. Contrasting ecological implications of food limita- tion in sea urchins and herbivorous gastropods. Mar. Ecol. Prog. Ser. 51:189-193. Bernard. F. R. & D, C. Miller. 1973. Preliminary investigation on the red sea urchin resources of British Columbia [Strongylocenlroliis fran- ciscanus (Ag.)|. Eish. Re.\. Bd. Can. Man. Rep. 1256:97 p. Black. R. C. Codd. D. Hebbert. S. Vink & J. Burt. 1984. The functional significance of the relative size of Aristotle's lantern in the sea urchin Echinomeira mathaei (de Blainville). J. E.xp. Mar. Biol. Ecol. 77:81- 97. PCOG & NSA. Lake Chelan, Washington AhMnicls. September 22-24. 1996 783 CONTENTS J. Harold Bealtie Butter clam [Sii.xidomus gigaiiteus) spawning trials 785 Jerry Bender The effects of mowing Spartina spp. during December and January in north Puget Sound as a non-chemical control methodology 7°5 A. Bradbury, W. A. Palsson and R. E. Pacunski Stock assessment of the commercial sea cucumber Paras!uiuipiis califoniuus in the San Juan Islands, Washington State, USA ''SS William W. Campbell and Jennifer A. Cahalan WDFW intertidal bivalve population assessment ui Puget Sound and Hood Canal, Washington 786 Jerry Chang and Kenneth K. Chew Positioning your shellfish in China 786 Anita E. Cook Intertidal shellfish management on public tidelands in Puget Sound. Washington — the management process 786 Annette Hoffmann. Jack V. Tagart and Dan Ayres Patterns in razor clam abundance estimates in coastal Washington 786 Manfred T. Kittel and Kenneth K. Chew Tasmanian Pacific oysters, Crassostrea gigas. in Washington state: characterization of a transplanted population 787 C.J. Langdon Update on the molluscan brookstock program — production of families 787 Xin Liu and A. M. Robinson Impact of cryoprotectants dimethyl sulfoxide, ethylene glycol, methanol, glycerol, sucrose and polyvinylpyrrolidone on oyster (Crassostrea gigas) eggs before freezing 787 Terrie A. Manning and Jennifer A. Cahalan Growth of the Pacific oyster. Crassostrea gigas (Thunberg) at 18 sites in Puget Sound and Hood Canal 787 Karl W. Mueller and Annette Hoffmann Effect of freshwater immersion on attachment in the Japanese oyster drill, Ceratostoma inornatum: implications for shellfish transfers in Washington state 788 Anja M. Robinson and Xin Liu Conservation of commercial Kumamoto oyster broodstock 788 Ervin J. Schumacker, Brett R. Dumbautd and Bruce E. Kauffman Preliminary results of investigations using oyster condition index to monitor the aquatic environment of Willapa Bay. Washington 788 David A. Sterritt. Elisabeth A. Wood and Jennifer L. Whitney Intertidal shellfish harvest assessment in Puget Sound. Washington 789 Kelly Toy Factors governing the distribution abundance, growth and reproduction of the freshwater mussel. Marguntifcra falcata, in forested watersheds of western Washington 789 William A. Wood Intertidal bivalve management on public tidelands in Puget Sound — an overview 789 786 Abstracls, September 22-24, 1996 PCOG & NSA, Lake Chelan, Washington omy, increased sample size, and the ability to survey several spe- cies simultaneously. WDFW INTERTIDAL BIVALVE POPULATION ASSESS- MENT IN PUGET SOUND AND HOOD CANAL, WASH- INGTON. William W. Campbell and Jennifer A. Cahalan, Washington Department of Fish and Wildlife. Point Whitney Shellfish Laboratory. 1000 Point Whitney Road. Brinnon. WA 98320. The intertidal shellfish project is responsible for determining clam and oyster populations on selected beaches in Hood Canal and Puget Sound. Clam and oyster population assessments have been conducted on Puget Sound beaches since the mid-1970s. Current survey methods are modified from methods developed in 1986. Approximately 30 to 50 clam and 20 to 30 oyster population assessments are conducted each season. Surveys are conducted on a beach specific basis using a sys- tematic random sampling design. Clam and oyster surveys are similar in nature, differing only in sample collection methods. Surveys are conducted on days with minus tides on 1 .0 ft MLLW or lower, for two hours on either side of low tide. The density of the resource and area of the population are estimated from these data, and population totals are calculated. Length-frequency in- formation is also collected during the surveys. Total allowable catch estimates are derived from the size-frequency distribution and population totals. The allowable catch is then applied to the following harvest season. New survey methods are currently being developed, incorpo- rating global positioning system (GPS) technology to determine individual sample locations. These sample locations will be used in conjunction with geographic information system (GIS. Maplnfo) to produce maps of beaches and the surveyed resources. In addition, significant landmarks, enhancement plots, and re- source density gradients will also be portrayed. KEY WORDS: Resource assessment, intertidal bivalves, GPS/ GIS POSITIONING YOUR SHELLFISH IN CHINA. Jerry Chang* and Kenneth K. Chew, School of Fisheries, Box 357980, University of Washington, Seattle, WA 98195-7980. The fast economic growth in mainland China demands more and more shellfish products from USA, particularly the North- west. Oysters, clams, mussels, lobsters and geoducks are the pri- mary products exported in alive, frozen, smoke and canned forms. Presently there is no well-established channel for shellfish growers to market their products in China. There are also some constraints challenging shellfish exporters, such as under-developed markets, poor infrastructure, inefficient brokers and agents, inadequate dis- tribution system as well as tariff and culture barriers to do business with China. Comments will reflect upon the present status of the seafood industry in China and what can be expected in the near and distant future. INTERTIDAL SHELLFISH MANAGEMENT ON PUBLIC TIDELANDS IN PUGET SOUND. WASHINGTON— THE MANAGEMENT PROCESS. Anita E. Cook, Washington De partment of Fish and Wildlife, Point Whitney Shellfish Labora- tory, 1000 Point Whitney Road, Brinnon, WA 98320. intertidal clam and oyster harvest comprises one of the largest recreational shellfish fisheries in Washington and provides har- vesting opportunities to thousands of harvesters each year. This paper describes current management techniques and challenges in the wake of the federal court decision regarding tribal shellfish rights. The intertidal shellfish management unit is primarily responsi- ble for the development of regional management plans and estab- lishing and implementing shellfish harvest regulations. The focus of this presentation is to elaborate on the process involved in developing fishery regulations and to discuss how recreational harvest estimates and population assessment information are inte- grated into management decisions. Consideration will be given to: I) development of regional management plans, 2) how estimates of recreational allowable catch are made within a cooperative state/tribal management planning process, 3) how pre-season pro- jections are developed, 4) inseason management, including season adjustments and harvest reporting, and 5) special challenges, in- cluding mixed species management, mixed fisheries (commercial and recreational), and information dissemination. KEY WORDS: Resource management, regulations, tribal, clams, oysters PATTERNS IN RAZOR CLAM ABUNDANCE ESTIMATES IN COASTAL WASHINGTON. Annette Hoffmann,* Jacli V. Tagart, and Dan Ayres, Washington Department of Fish and Wildlife. Washington. 98501. Precise estimates of razor clam {Siliqua patula) abundance are necessary for responsible harvest management and for monitoring of long term population trends. Understanding "patchiness." or how the patterns of densities are distributed can greatly help one improve the precision of an abundance estimator by stratifying the beach into areas of like densities. Stratification will improve the precision of an estimator if the within strata variance is smaller, on average, than the among strata variance. However, stratification will degrade the precision if the within variance is the same or greater than the among variance, or if the process of stratifying intensifies existing biases. If stratification variables can be iden- tified that are consistent across beaches and persistent across years, then we will use them in our abundance monitoring and harvest management plans. We investigated the use of stratifying razor clam densities along elevational and latitudinal (transects perpendicular to the surf) gradients on three beaches using pump survey data collected from 1994-1996. Analysis results showed that elevation was a good stratification variable, but that latitude was not. However, a secondary study showed that the "patchiness" of razor clam dis- PCOG & NSA, Lake Chelan, Washington Abstrmlx. September 22-24, 1996 787 tribution was on a smaller scale than the latitudinal transects and therefore, would be unlikely to show up as a latitudinal effect. Stratification by elevation resulted in improved precision in 5 out of 6 beach-years. The case where elevation was not significant occurred in a year of relatively low abundance. In this case we also encountered both kinds of problems one can experience with strat- ification; intensified bias and greater within than among strata variances. With these data, we cannot discern the cause of the non-significance. These results suggest that elevation may be a consistent strat- ification variable, but may not be persistent. However, these re- sults are preliminary We will further investigate both variables on more beaches and over more years before concluding that they should or should not be used in a monitoring plan. TASMANIAN PACIFIC OYSTERS, CRASSOSTREA GIGAS, IN WASHINGTON STATE: CHARACTERIZATION OF A TRANSPLANTED POPULATION. Manfred T. Kittel* and Kenneth K. Chew, School of Fisheries. Box 357980. University of Washington. Seattle. WA 98195. In 1995. 32 Pacific oysters from Tasmania. Australia, were artificially spawned in quarantine in the state of Washington and the resulting F, generation was outplanted at three different loca- tions of Puget Sound. We have begun studies to characterize these oysters at several levels. Survival, growth rates and shell morphol- ogy of oysters grown under different environmental regimes are determined and compared to similar data from C. gifius of local ongin. Patterns of gametogenesis and glycogen storage will be examined from histological sections of oysters taken over the pe- riod of one growing season. The ability of the imported oysters to form viable hybrids with C. gigas of local origin and closely related species such as C. sikamea will be determined from recip- rocal crosses between selected oysters. Molecular analysis of the transplanted oysters will focus on establishing their species iden- tity and genetic relatedness to local C. gigas populations. Prelim- inary findings of some of the ongoing studies will be presented. UPDATE ON THE MOLLUSCAN BROODSTOCK PRO- GRAM—PRODUCTION OF FAMILIES. Chris J. Langdon,* Hatfield Marine Science Center, Oregon State University, New- port, OR 97365. The focus of the Molluscan Broodstock Program (MBP) during the first year of funding has been rearing two groups of about 50 families of Pacific oyster spat for planting and evaluation at com- mercial test sites on the West Coast, U.S. Pairs of oysters from either Willapa or Dabob Bay "wild" populations were crossed to produce full-sib families. Larvae were reared in UK) 1 tanks at a concentration of 4 larvae/ml. Competent larvae were exposed for 2 h to 2 x lO"* M epinephrine to induce metamorphosis. Throughout the culture period, larvae were fed on a mixed algal diet of Chaeloceros sp. and a flagellate (Pseudo- isochrysis or Isochrysis sp.). Spat were initially cultured in a small (MINI) upweller system at 25°C and fed on the mixed algal diet. When all the spat from a family had been collected, the number of spat per family was reduced to 20.000 by random partitioning. When spat could be retained on a 1.5 mm mesh, they were transferred to a larger (MAXI) upweller system and cultured at 2(>-25°C on a diet con- sisting of mainly Chaeloceros sp. When spat in the MAXI system could be retained on a 'A inch mesh, they were transferred to '/s inch mesh bags suspended in Yaquina Bay, Oregon, and held for planting at commercial test sites in late summer/fall 1996. Survival, growth and meat yields of planted families will be compared when they reach market size and top performing fami- lies will be used to produce the next MBP generation. IMPACT OF CRYOPROTECTANTS DIMETHYL SULF- OXIDE, ETHYLENE GLYCOL. METHANOL, GLYC- EROL, SUCROSE AND POLYVINYLPYRROLIDONE ON OYSTER {CRASSOSTREA GIGAS) EGGS BEFORE FREEZ- ING. Xin Liu* and A. M. Robinson, Department of Fisheries and Wildlife, Hatfield Marine Science Center, Oregon State Uni- versity, Newport. OR 97365. As a basis for Pacific oyster cryopreservation study, informa- tion on impact of cryoprotectants on oyster eggs before freezing is an important factor. Oyster eggs were exposed to various concen- trations of six cryoprotective compounds, dimethyl sulfoxide (DMSO), ethylene glycol (EG), methanol, glycerol, sucrose and polyvinylpyrrolidone (PVP) at room temperature (21-24°C) for 30 minutes. The results showed that glycerol at tested concentrations and methanol at greater than 1.2 M concentrations were highly toxic to oyster eggs, whereas sucrose at tested concentrations and PVP at less than 10% concentrations did not have toxic effects. The results of time exposure experiment conducted with 1 .4 M DMSO. 1.8 M EG. 2,4 M methanol, 1.4 M glycerol, 0.292 M sucrose and 10% PVP concentrations for 25 minutes indicated that the exposure time should be less than 20 minutes to minimize the injury of the eggs caused by cryoprotectants. The combination of 0.7 M DMSO -I- 0.141 M sucrose concentration mixture (15 minute exposure) improved oyster egg survival rate. It could be attributed to the replacement of DMSO fractions with sucrose rather than any specific toxicity blocking mechanism. Exposure to the cryoprotectants before freezing can cause major biochemical or/and osmotic injury on oyster eggs. GROWTH OF THE PACIFIC OYSTER, CRASSOSTREA GI- GAS (THUNBERG) AT 18 SITES IN PUGET SOUND AND HOOD CANAL. Terrie A. Manning and Jennifer A. Cahalan, Department of Fish and Wildlife, Pt. Whitney Shellfish Labora- tory, 1000 Pt. Whitney Rd.. Brinnon. WA 98320. A mark and recapture study was conducted at 18 beaches in Hood Canal and Puget Sound, Washington to obtain accurate es- timates of Pacific oyster growth rates. These population parameter THE NATIONAL SHELLFISHERIES ASSOCIATION The National Shellfisheries Association (NSA) is an international organization of scientists, manage- ment officials and members of industry that is deeply concerned and dedicated to the formulation of ideas and promotion of knowledge pertinent to the biology, ecology, production, economics and man- agement of shellfish resources. The Association has a membership of more than 1000 from all parts of the USA. Canada and 18 other nations: the Association strongly encourages graduate students' mem- bership and participation. WHAT DOES IT DO? — Sponsors an annual scientific conference. — Publishes the peer-reviewed Jdurnal of Shellfish Rcsfarch. — Produces a Quarterly Newsletter. — Interacts with other associations and industry. WHAT CAN IT DO FOR YOU? — You will meet kindred scientists, managers and industry officials at annual meetings. — You will get peer review through presentation of papers at the annual meeting. — If you are young, you will benefit from the experience of your elders. — If you are an elder, you will be rejuvenated by the fresh ideas of youth. — If you are a student, you will make useful contacts for your job search. — If you are a potential employer, you will meet promising young people. — You will receive a scientific journal containing important research articles. — You will receive a Quarterly Newsletter providing information on the Association and its activities, a book review section, information on other societies and their meetings, a job placement section, etc. HOW TO JOIN — Fill out and mail a copy of the application blank below. The dues are 45 US $ per year ($25 for students) and that includes the Joiinud and the Newsletter! NATIONAL SHELLFISHERIES ASSOCIATION— APPLICATION FOR MEMBERSHIP {NEW MEMBERS ONLY) Name: For the calendar year: Date: Mailing address: Institutional affiliation, if any: Shellfishery interests: Regular or student membership: Student members only — advisor's signature REQUIRED: Make cheques {MUST be drawn on a US bank), international postal money orders or VISA for $45 ($25 for students with advisor's signature) payable to the National Shellfisheries Association and send to Christine Hodgson, BC Ministry of Ag. and Fisheries, 2500 Cliffe Avenue. Courtenay, British Columbia, CANADA V9R 5M6. yjhS .i3 Cameron P. Goater and Amy W. Weber Factors affecting the distribution and abundance of Mytilicola orientalis (Copepoda) in the mussel, Mytilus trossuhis. in Barkley Sound, B.C 681 Paul A. Haefner, Jr., Becky Sheppard, Julie Barto, Erin McNeil and Vanessa Cappellino Application of ultrasound technology to molluscan physiology: noninvasive monitoring of cardiac rate in the blue mussel, Mytilus eduUs Linnaeus, 1758 685 Arnold G. Eversole, Christopher J. Kempton, Nancy H. Hadley and William R. Buzzi Comparison of growth, survival, and reproductive success of diploid and triploid Mercenaria mercenaria 689 Stephen J. Kleinschuster, Jason Parent, Charles W. Walker and C. Austin Farley A cardiac cell line from Mya areiiaria (Linnaeus, 1759) 695 Dan C. Marelli and William S. Arnold Growth and mortality of transplanted juvenile hard clams, Mercenaria mercenaria. in the northern Indian River lagoon, Florida ™" Dorset H. Hurley and Randal L. Walker The effects of larval stocking density on growth, survival, and development of laboratory-reared Spisula soUdissima similis (Say. 1822) 715 H.-Jorg Urban Population dynamics of the bivalves Venus antiqua. Tagelus dombeii. and Ensis macha from Chile at 36°S 719 Gudrun G. Thorarinsdottir and Gardar Johannesson Shell length-meat weight relationships of ocean quahog, Arctica islandica (Linnaeus, 1767), from Icelandic waters ... 729 Lincoln Mackenzie, David White and Janet Adamson Temporal variation and tissue localization of paralytic shellfish toxins in the New Zealand tuatua (surfclam), Paphies suhlriangulala ' ^' Marcial Villalejo-Fuerte, Bertha Patricia Ceballos-Vazquez and Federico Garcia-Dominguez Reproductive cycle of Laevicardium elalum (Sowerby, 1833) (Bivalvia: Cardiidae) in Bahia Concepcion, Baja California Sur, Mexico 741 Joseph Biss HI, Franck H. Laruelle and Daniel P. Molloy Use of sieves for the rapid size selection of Dreissena polymorpha samples 747 George R. Abbe and Cluney Stagg Trends in blue crab (CalUnectes sapidus Rathbun) catches near Calvert Cliffs, Maryland, from 1968 to 1995 and their relationship to the Maryland commercial fishery 75 1 Hui Liu and James W. Avault, Jr. Effect of nitrite on growth of juvenile red swamp crawfish, Procambarus clarkii 759 Donna G. Ingle and Robert M. Ingle Notes on the blue crab fishery in the Apalachicola, Florida estuary 763 Elizabeth Wenner, H. Randall Beatty and Loren Coen A method for quantitatively sampling nekton on intertidal oyster reefs 769 T. J. Morris and A. Campbell Growth of juvenile red sea urchins (Strongyiocentrotus franciscanus) fed Zostera marina or Nereocystis luelkeana — 777 Abstracts of technical papers presented at the 50th Annual Meeting of the Pacific Coast Oyster Growers Association and National Shellfisheries Association, September 22-24, 1996, Lake Chelan, Washington 781 COVER PHOTO: Peari oyster collectors in the waters in front of Nusa Tupe, ICLARM's Fieldstation in the Western Province of Solomon Islands. (Photo by Mike McKoy) The Journal of Shellfish Research is indexed in the following: Science Citation Index®, Sci Search®, Research Alert®. Current Contents®/ Agriculture, Biology and Environmental Sciences, Biological Abstracts, Chemical Abstracts, Nutrition Abstracts, Current Advances in Ecological Sciences, Deep Sea Research and Oceanographic Literature Review, Environmental Periodicals Bibliography, Aquatic Sciences and Fisheries Abstracts, and Oceanic Abstracts. JOURNAL OF SHELLFISH RESEARCH Vol. 15, No. 3 DECEMBER 1996 CONTENTS Gary H. Wikfors Honored Life Member; Robert R . L. Guillard 533 Kim J.Friedman and Johann D. Bell Effects of different substrata and protective mesh bags on collection of spat of the pearl oysters. Pinctada margaritifera (Linnaeus, 1758) and Pinctada maculata (Gould, 1850) 535 William S. Fisher, James T. Winstead, Leah M. Oliver, H. Lee Edmislon and George O. Bailey Physiologic variability of eastern oysters from Apalachicola Bay, Florida 543 William S. Fisher, Leah M. Oliver and Patrice Edwards Hematologic and serologic variability of eastern oysters from Apalachicola Bay, Florida 555 Herbert M. Austin, David Evans and Dexter S. Haven A retrospective time series analysis of oyster, Crassostrea virginica. recruitment ( 1946-1993) 565 Diane J. Brousseau Epizootiology of the parasite, Perkinsus maiiniis (Dermo) in intertidal oyster populations from Long Island Sound. .. . 583 Emily Y . Chang, Steven L. Coon, Marianne Walch and Ronald Weiner Effects oi Hyphomoiias PM-1 biofilms on the toxicity of copper and zinc to Crassostrea gigas and Crassostrea virginica larval set 589 Carolyn S. Friedman Haplosporidian infections of the Pacific oyster, Crassostrea gigas (Thunberg), in California and Japan 597 Oscar Arizpe C. Secondary production, growth and survival of the Pacific oyster Crassostrea gigas (Thunberg) in tropical waters, Bahia de La Paz, Mexico 601 Caleb Gardner, Greg B. Maguire and Greg N. Kent Studies on triploid oysters in Australia. VIL Assessment of two methods for determining triploidy in oysters: adductor muscle diameter and nuclear size 609 A. G. Jeffs, R. G. Creese and S. H. Hooker Annual pattern of brooding in populations of Chilean oysters, Tiostrea chilensis. (Philippi, 1845) from northern New Zealand 617 Gisele Signoret-Brailovsky , Alfonso N. Maeda-Martinez, Teodoro Reynoso-Granados, Ernesto Soto-Galera, Pablo Monsalvo-Spencer and Gabriela Valle-Meza Salinity tolerance of the catarina scallop Argopecten ventricosiis-circiilaris (Sowerby II, 1835) 623 Michael P. Heasman, Wayne -4. O'Connor and Allen W. J. Frazer Ontogenetic changes in optimal rearing temperatures for the commercial scallop. Pectcn fuinatus Reeve 627 Yantian T. Lu and Norman J. Blake Optimum concentrations oi Isochrysis galhana for growth of larval and juvenile bay scallops, Argopecten irraduins concentriciis ( Say ) 635 C. Rios, J. Canales and J. B. Peria Genotype-dependent spawning; evidence from a wild population of Pecten jacohaeus (L.) (Bivalvia: Pectinidae) 645 N. K. Stepto and P. A. Cook Feeding preferences of the juvenile South African abalone Haiiotis inidac (Linnaeus. 1758) 653 G. P. Hawkes, R. W. Day, M. W. Wallace, K. W. Nugent, A. A. Bettiol, D. N. Jamieson and M. C. Williams Analyzing the growth and form of mollusc shell layers, in situ, by cathodoluminescence microscopy and Raman spectroscopy 659 Jorge Caceres-Martinez, Rebeca Vasquez-Yeomans and Eduardo Suarez Morales Two parasitic copepods, Pseudomyicola spinosus and Modiolicola gracilis, associated with edible mussels, Mytilus galloprorincialis and Mytilus californianus. from Baja California, NW Mexico 667 J. I. P. Iglesias, A. Perez Camacho, E. Navarro, U. Labarta, R. Beiras, A. J. S. Hawkins and J. Widdows Microgeographic variability in feeding, absorption, and condition of mussels (Mytilus galloprarincialis Lmk): a transplant experiment ^73 CONTENTS CONTINUED ON INSIDE BACK COVER lipli